* fix(inbox): don't archive after deleting an issue
Deleting an issue from the Inbox page was calling the archive API on the
inbox item right after deleteIssue succeeded. Because the inbox_item row
has ON DELETE CASCADE on issue_id, it was already gone by then and the
archive call 404'd with "inbox item not found", surfacing a "Failed to
archive" toast.
Drop the redundant archive call and invalidate the inbox cache through
the issue:deleted WS handler so every tab stays in sync without an extra
round trip.
* fix(inbox): keep stale selection on /inbox instead of the deleted issue
When another tab deletes the selected inbox issue, onInboxIssueDeleted
prunes the cache and `selected` becomes null. The existing fallback then
redirected to the issue detail page — which is also gone, so the user
landed on a "This issue does not exist..." screen instead of back in the
inbox list.
Track the last key that actually resolved against the inbox list. If it
used to be in the list and just disappeared, clear the selection and
stay on /inbox. Only shared links that were never in the user's inbox
continue to fall back to the issue detail page.
Also add ws-updaters tests covering onInboxIssueDeleted and
onInboxIssueStatusChanged.
* feat(issues): paginate every status column, not just done
Previously the workspace issues list fetched all non-done/cancelled
issues in a single unbounded `open_only=true` request and only
paginated the done column. In workspaces with many open issues this
ballooned the initial payload and skipped pagination entirely.
Restructure the issue list cache into per-status buckets
(`{ byStatus: { [status]: { issues, total } } }`) fetched in parallel,
generalize `useLoadMoreDoneIssues` into `useLoadMoreByStatus(status,
myIssuesOpts?)`, and render an infinite-scroll sentinel inside every
accordion group and kanban column. Sort and filter stay client-side,
matching the done column's existing behavior.
Backend `ListIssues` already supports per-status pagination, so no
API changes are required.
* fix(issues): handle project / hidden-column / lookup regressions from paginated list cache
After bucketing the issue list cache by status, three consumers that
treated `issueListOptions()` as a complete local index broke:
- `project-detail.tsx` filtered the workspace list by `project_id`
client-side, so projects whose issues sat past the first 50-per-status
page rendered empty. Switch to `myIssueListOptions(wsId,
'project:<id>', { project_id })` so the server returns only this
project's issues; add `project_id` to `ListIssuesParams` /
`MyIssuesFilter` / api client.
- `board-view.tsx` HiddenColumnsPanel read counts from the in-memory
`issues` array — a paginated fragment. Pass `myIssuesOpts` through to
a per-row subcomponent that reads the real per-status total from the
cache.
- `tasks-tab.tsx` and `search-command.tsx` used the list as a global
lookup for task titles / Recent items / current-issue chrome. Switch
both to per-id `issueDetailOptions` via `useQueries` so they're
independent of which page the issue lands on.
Drop the now-redundant `doneTotal` override prop on BoardView/ListView
and the `allIssues` prop on BoardView (only HiddenColumnsPanel consumed
it).
Tests updated: tasks-tab now mocks `api.getIssue`; search-command mocks
`issueDetailOptions` + `useQueries`; project-issue-metrics drops the
`doneColumnCount` assertion.
Two unrelated bugs were preventing the GitHub-hosted runner desktop
release matrix from succeeding:
1. Windows job failed with `spawnSync electron-vite ENOENT`. On
Windows the package-local binaries are `.cmd` shims and Node's
`spawnSync` does not consult PATHEXT unless going through a shell.
Pass `shell: true` for both the electron-vite and electron-builder
spawns; on POSIX hosts these are real executables so the shell hop
is harmless.
2. Linux `.deb`/`.rpm` job failed with electron-builder errors:
`Please specify project homepage` and `Please specify author
'email'`. fpm requires a maintainer when generating .deb, and
electron-builder derives it from the app package.json metadata. Add
`description`, `homepage`, `repository`, `author` (with
email) and `license` to apps/desktop/package.json so the Linux
targets have the metadata they need.
Refs: https://nodejs.org/api/child_process.html#spawning-bat-and-cmd-files-on-windows
Refs: https://www.electron.build/configuration.html#metadata
Co-authored-by: Eve <eve@multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Agents can end a comment-triggered run without calling `multica issue comment
add` — the final reply stays in terminal / run-log text and never reaches
the user, even though the run panel shows "Completed". PR #1372 addressed
this via prompt wording, but compliance is inherently best-effort.
The server already had an exact fix for the assignment-triggered branch:
`HasAgentCommentedSince` + fallback synthesis from `payload.Output`. The
comment-triggered branch was explicitly exempted on the theory that the
agent "replies via CLI with --parent, so posting here would create a
duplicate" — but that is precisely the path that's failing.
Remove the `!task.TriggerCommentID.Valid` guard so the invariant "every
completed issue task has at least one agent comment on the issue" holds for
both branches. The existing `HasAgentCommentedSince` check still prevents
duplicates for compliant agents, and `createAgentComment` already threads
the synthesized comment under `task.TriggerCommentID` when present.
Regression tests cover both:
- comment-triggered + silent agent → synthesized comment threaded under trigger
- comment-triggered + agent already posted → no duplicate
The renderer's navigation adapter fell back to https://multica.ai when
VITE_APP_URL was unset (i.e. desktop dev builds), so "Copy link" in a dev
build produced a production URL instead of one pointing at the running
dev web frontend. Match the fallback used by pages/login.tsx
(http://localhost:3000) so dev links stay on the dev host.
The inbox detail panel keyed `<IssueDetail>` by `selected.id` (inbox-item
id). `deduplicateInboxItems` picks the most recent inbox notification per
issue, so every new `comment:created` / `reaction:added` event for the
currently open issue produced a fresh inbox item with a new id — flipping
the React key and forcing a full unmount/remount of `IssueDetail`. That
wiped the comment composer draft, dropped focus, and reset scroll.
Key on `selected.issue_id` instead: stable for the life of an open issue
(so input + scroll survive incoming events) and still changes when the
user picks a different issue (so state resets between issues, as before).
The bluemonday HTML sanitizer applied to comment content (added in #679)
treats Markdown source as HTML, entity-encoding syntactically meaningful
characters and normalizing whitespace. This corrupts user input:
- "> quote" -> "> quote" (blockquote lost, see #1303)
- '"foo"' -> '"foo"' (literal entities visible)
- "\n\n2." -> " 2." (ordered list items merged into prose)
Comment content is stored as Markdown source. XSS is already handled at
two layers:
- Render: rehype-sanitize in packages/ui/markdown and
packages/views/editor/readonly-content (mention:// allowlist,
data-href restricted to http(s), class restricted to
code/div/span/pre).
- Edit: @tiptap/markdown is configured with html:false, so Markdown
source containing raw HTML tags is treated as plain text.
Removing the server-side sanitizer therefore does not lower the security
boundary, and restores faithful Markdown round-tripping.
The PR #1342 workaround in the editor serializer can be dropped once
this lands.
Co-authored-by: devv-eve <eve@devv.ai>
Co-authored-by: Eve <eve@multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
posthog-js ships with autocapture, heatmaps, dead-click detection,
session recording, exception capture, and surveys all on by default.
Staging verification showed the Activity view flooded with "clicked
button" / "clicked span with text \"…\"" events — they leak
user-typed content into PostHog, burn the billed event budget, and
dilute the explicit funnel. Our product analytics surface is narrow
and intentional (see docs/analytics.md): only the events we emit
server-side plus one manual $pageview belong. Opt all the auto
surfaces off at init time so the Activity view reflects the funnel.
* feat(analytics): add PostHog client with async batch shipping
Introduces server/internal/analytics, the shipping layer for the product
funnel defined in docs/analytics.md. Capture is non-blocking — events are
enqueued into a bounded channel and a background worker batches them to
PostHog's /batch/ endpoint. A broken backend drops events rather than
blocking request handlers.
Local dev and self-hosted instances run a noop client until the operator
sets POSTHOG_API_KEY. This is PR 1 of MUL-1122; signup and workspace_created
emission land in the follow-up commit so this change is independently
reviewable.
* feat(server): emit signup and workspace_created analytics events
Wires analytics.Client through handler.New and main, then emits the first
two funnel events:
- signup fires from findOrCreateUser (which now reports isNew), covering
both the verification-code and Google OAuth entry points — a single
emission site guarantees Google signups aren't missed.
- workspace_created fires after the CreateWorkspace transaction commits,
with is_first_workspace computed from a post-commit ListWorkspaces count
so we can distinguish fresh-user activation from returning-user
expansion.
Tests use analytics.NoopClient so nothing ships from test runs. PR 1 of
MUL-1122; runtime_registered and issue_executed follow in later PRs per
the plan.
* refactor(analytics): drop is_first_workspace from workspace_created
Stamping "is this the user's first workspace?" at emit time races under
concurrent CreateWorkspace requests: two transactions committing close
together can both read a post-commit count greater than one and both emit
false. Fixing it at the SQL layer requires a schema change we don't want in
PR 1.
PostHog answers the same question exactly from the event stream (funnel on
"first time user does X" / cohort on $initial_event), so removing the
property loses no information and makes the emit side race-free.
* docs(analytics): document self-host safety defaults
Spell out why self-hosted instances never ship events upstream by default
(empty POSTHOG_API_KEY → noop client) and explain how operators can point
at their own PostHog project without any code change.
* feat(analytics): emit runtime_registered, issue_executed, team_invite_*
Three server-side funnel events, all gated on first-time state transitions
so retries and re-runs don't inflate the WAW buckets:
- runtime_registered fires from DaemonRegister when UpsertAgentRuntime
reports (xmax = 0) — i.e. the row was inserted, not updated. Heartbeats
and re-registrations stay silent.
- issue_executed fires from CompleteTask after an atomic
UPDATE issue SET first_executed_at = now() WHERE id = $1 AND
first_executed_at IS NULL flips the column for the first time. Retries,
re-assignments, and comment-triggered follow-up tasks hit the WHERE
clause and no-op. Carries nth_issue_for_workspace so the ≥1/≥2/≥5/≥10
buckets filter without extra queries.
- team_invite_sent fires from CreateInvitation and team_invite_accepted
from AcceptInvitation, closing the expansion funnel.
Adds a 050 migration for issue.first_executed_at plus a partial index so
the workspace-scoped executed-count query doesn't scan the never-executed
tail.
* feat(config): surface PostHog key via /api/config
Extends AppConfig with posthog_key / posthog_host sourced from env on
every request (so operators can rotate the key via secret refresh without
a restart). Reading the key off the server — rather than baking it into
the frontend bundle via NEXT_PUBLIC_* — means self-hosted instances
inherit the blank key automatically and never ship events upstream.
* feat(analytics): wire posthog-js identify + UTM capture on the client
Adds @multica/core/analytics — a thin wrapper around posthog-js that owns
attribution capture and identity merge. Posthog-js config comes from
/api/config (not NEXT_PUBLIC_*), so self-hosted instances whose server
returns an empty key automatically run the SDK inert.
captureSignupSource stamps a multica_signup_source cookie with UTM params
and the referrer's origin (never the full referrer — that can leak OAuth
code/state in the callback URL). The backend signup event reads this
cookie on new-user creation.
Identity flows:
- auth-initializer fires identify() right after getMe() resolves, on both
cookie and token paths. A getConfig/getMe race is handled by buffering
a pending identify inside the analytics module and flushing it once
initAnalytics finishes.
- auth store calls identify() on verifyCode / loginWithGoogle /
loginWithToken and resetAnalytics() on logout so the next login merges
cleanly without bleeding events.
* docs(analytics): describe runtime_registered, issue_executed, invite events
Fills in the schema for the remaining funnel events. Captures the
design commentary that belongs next to the contract rather than in a PR
description — in particular why issue_executed uses the atomic
first_executed_at flip instead of counting task-terminal events, and why
runtime_registered relies on xmax = 0 rather than a query-then-write.
* fix(analytics): drop non-atomic nth_issue_for_workspace from issue_executed
Computing the workspace's Nth-issue ordinal at emit time is not atomic
under concurrent first-completions — two transactions can both run
MarkIssueFirstExecuted, then both run CountExecutedIssuesInWorkspace, and
both observe count=1 before either has committed, so both events go out
stamped as n=1. Serialising it would mean a per-workspace advisory lock
or a SERIALIZABLE-isolated tx; PostHog answers the same question exactly
at query time via row_number() partitioned by workspace_id, so the
emit-time property adds risk without adding information.
Removes the property from analytics.IssueExecuted, deletes the unused
CountExecutedIssuesInWorkspace query, and regenerates sqlc. The partial
index stays — any future workspace-scoped executed-issue query will want
it.
* fix(analytics): wire $pageview and harden signup_source cookie payload
Two frontend fixes from the PR review:
- PageviewTracker, mounted under WebProviders, fires capturePageview on
every Next.js App Router path / query-string change. Without this the
capturePageview helper in @multica/core/analytics was never called and
the acquisition funnel's / → signup step was empty.
- captureSignupSource now caps each UTM / referrer value at 96 chars
*before* JSON.stringify, and drops the whole cookie when the serialised
payload still exceeds 512 chars. Previously the overall slice(0, 256)
could leave a half-JSON string on the wire that neither the backend nor
PostHog could parse.
Both capturePageview and identify now buffer a single pending call when
fired before initAnalytics resolves — otherwise the initial "/" pageview
and same-turn login identify race the /api/config fetch and get dropped.
resetAnalytics clears both buffers so a logout→login cycle stays clean.
* fix(analytics): URL-decode signup_source cookie on read
Go does not URL-decode Cookie.Value automatically, so the frontend's
JSON-then-encodeURIComponent payload was landing in PostHog as
percent-encoded garbage (%7B%22utm_source...). Unescape on read so the
backend receives the original JSON string the frontend intended, and
drop values that fail to decode or exceed the server-side cap — sending
truncated garbage is worse than sending nothing. Oversized-cookie guard
matches the frontend's SIGNUP_SOURCE_MAX_LEN.
* docs(analytics): reflect nth-issue drop, $pageview wiring, cookie encoding
Pulls the schema doc back in line with the code: issue_executed no longer
advertises nth_issue_for_workspace (with a note about why PostHog derives
it at query time instead), the frontend $pageview section names the
actual PageviewTracker component that fires it, and the signup_source
section documents the per-value cap / overall drop rule and the
encode-on-write / decode-on-read contract.
---------
Co-authored-by: Jiang Bohan <bhjiang@outlook.com>
* feat(desktop): support macOS cross-platform packaging
* fix(desktop): use releaseType instead of publishingType in electron-builder publish config
publishingType is not a valid electron-builder key; the correct GitHub
provider option is releaseType. The previous value was silently ignored,
causing uploads to be skipped and breaking auto-update.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(release): standardize artifact naming across desktop and CLI
Unified scheme: `multica-<kind>-<version>-<platform>-<arch>.<ext>` so a
filename alone reveals kind, version, platform, and CPU arch.
Desktop (apps/desktop/electron-builder.yml):
mac → multica-desktop-<v>-mac-<arch>.{dmg,zip}
linux → multica-desktop-<v>-linux-<arch>.{deb,AppImage}
(fixes `\${name}` expanding the scoped `@multica/desktop` into a
broken `@multica/desktop-*` filename path)
windows → multica-desktop-<v>-windows-<arch>.exe
CLI (.goreleaser.yml):
multica_<os>_<arch>.tar.gz → multica-cli-<v>-<os>-<arch>.tar.gz
(adds `-cli` marker + version; switches `_` to `-` for consistency)
Matrix update in apps/desktop/scripts/package.mjs `--all-platforms`:
- drop mac x64 (Intel not a target yet)
- add linux arm64
Final: mac arm64, win x64/arm64, linux x64/arm64.
Downstream updates so install paths match the new CLI names:
- scripts/install.sh
- scripts/install.ps1 (URL + checksum regex)
- CLI_INSTALL.md
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(release): use multica_{os}_{arch} CLI archive naming
Standardize on the GoReleaser default 'multica_{os}_{arch}.{tar.gz|zip}'
asset names. Install scripts and the desktop CLI bootstrap now resolve
assets via checksums.txt so they work without hardcoding versions.
The Go self-update path queries the GitHub release API and accepts
either the new or legacy 'multica-cli-<version>-...' names so existing
releases keep updating cleanly.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(release): ship both legacy and versioned CLI archive names
GoReleaser now produces both 'multica_{os}_{arch}.{ext}' (legacy) and
'multica-cli-{version}-{os}-{arch}.{ext}' (versioned) archives in every
release. The legacy name keeps already-released CLIs self-updating; the
versioned name is what new clients should use going forward.
Self-update / install paths flipped to prefer the versioned name and
fall back to legacy:
- server/internal/cli/update.go (multica update)
- apps/desktop/src/main/cli-release-asset.ts (desktop CLI bootstrap)
- scripts/install.sh, scripts/install.ps1 (fresh install)
Homebrew formula is pinned to the versioned archive via 'ids: [versioned]'.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(desktop): also build Linux .rpm packages
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(release): build Linux/Windows Desktop installers in CI; detect Windows ARM64 in install.ps1
Address review feedback on PR #1262:
- .github/workflows/release.yml: add a 'desktop' job that runs after the
CLI 'release' job and packages the Desktop installers for Linux
(AppImage/deb/rpm) and Windows (NSIS) on x64 and arm64, then publishes
them to the same GitHub Release via electron-builder. macOS Desktop
continues to ship through the manual release-desktop skill so it can
be signed and notarized with Apple Developer credentials.
- scripts/install.ps1: detect Windows ARM64 hosts via
RuntimeInformation::OSArchitecture so the new windows-arm64 CLI
archive is downloaded on ARM64 machines instead of always falling
back to amd64.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(release): split Windows arm64 auto-update channel to avoid latest.yml collision
electron-builder's update metadata file is hardcoded to `latest.yml` for
Windows regardless of arch (only Linux gets an arch-suffixed name; see
app-builder-lib's getArchPrefixForUpdateFile). With two separate
electron-builder invocations for Windows x64 and arm64, both publish
`latest.yml` to the same GitHub Release and the second upload silently
overwrites the first — leaving one of the two architectures with auto-
update metadata pointing at the other arch's installer.
Route Windows arm64 to its own `latest-arm64` channel:
* scripts/package.mjs appends `-c.publish.channel=latest-arm64` only
for the Windows arm64 invocation, so x64 keeps producing `latest.yml`
and arm64 produces `latest-arm64.yml` alongside it.
* updater.ts pins `autoUpdater.channel = 'latest-arm64'` on Windows
arm64 clients so they fetch the matching metadata file.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
OpenClaw's `--json` result blob carries the actual LLM identifier in
`meta.agentMeta.model` (e.g. `deepseek-chat`, `claude-sonnet-4`),
alongside `provider` and the usage breakdown. The backend was reading
the surrounding `agentMeta.usage` and `agentMeta.sessionId` but skipping
the `model` field entirely, then attributing every run's tokens to
`opts.Model` — which for openclaw is the *agent name* passed via
`--agent`, not a real model identifier — falling all the way through to
"unknown" when no agent.model was configured.
Surface the runtime-reported model:
- `openclawEventResult` gains a `model` string.
- `buildOpenclawEventResult` reads `agentMeta.model` (trimmed; empty
string when absent for forward-compat with older runtimes / partial
outputs).
- `processOutput` propagates it through the result-blob branch.
- `Execute`'s usage map prefers `scanResult.model`, falling back to
`opts.Model` then `"unknown"` — preserving the prior behavior path
for any runtime that doesn't surface its own model yet.
Two unit tests cover both the populated and missing cases.
Refs: #1395
Surface the actual exec path + argv for every agent backend at INFO
so operators can see the exact command without flipping to debug.
Also add the missing log line in pi.go for consistency with the
other nine backends.
* fix(cli): detach daemon from parent console on Windows
CREATE_NEW_PROCESS_GROUP alone leaves the daemon attached to the
parent console, so closing the launching cmd/PowerShell window fires
CTRL_CLOSE_EVENT down the inherited console and takes the daemon
with it. Add DETACHED_PROCESS so the child has no console at all;
stdout/stderr are already redirected to the log file before spawn.
* fix(cli): make `multica update` work while the binary is running on Windows
On Windows, a running .exe is opened without FILE_SHARE_WRITE, so the
previous os.Rename(tmp, exe) always failed with "Access is denied" —
every `multica update` on Windows hit this, because the CLI is
updating its own running binary.
Windows does allow renaming the running .exe (just not overwriting
it), so the new Windows-only replaceBinary moves the running binary
to `.old` first, installs the new one, and restores the original if
installation fails. A best-effort CleanupStaleUpdateArtifacts runs
at CLI/daemon startup to reclaim the leftover `.old` file once the
old process has exited.
Unix keeps the plain rename-over semantics (the old inode stays valid
for the running process).
* fix(cli): stop daemon via HTTP /shutdown instead of console ctrl events
With DETACHED_PROCESS the Windows daemon shares no console with the
stop caller, so `GenerateConsoleCtrlEvent(CTRL_BREAK_EVENT, pid)`
silently never reaches it — the old code would report "stop sent"
while the daemon kept running. Replace the platform-specific
stopDaemonProcess with a cross-platform POST to the daemon's HTTP
/shutdown endpoint, which cancels the same top-level context the
self-restart path already uses. Fall back to `process.Kill()` if
the HTTP call fails.
Also drops the now-unused stopDaemonProcess / CTRL_BREAK_EVENT
wiring, adds handler tests, and updates the DETACHED_PROCESS comment.
Extract the URL assembly at the end of S3Storage.Upload into a helper
(uploadedURL) so the four env-var combinations can be covered by a
table-driven test without mocking s3.PutObject. This locks in the fix
from #1300 — cdn > endpoint > bucket — so future refactors can't
silently regress the CDN-wins-over-custom-endpoint case.
No behavior change.
Phase 0 hotfix for the cross-workspace contamination reported in MUL-1027
/ #1235: an agent running for workspace A ended up commenting on (and
renaming) a two-day-old issue in workspace B.
#1249/#1259 fixed resolution for autopilot tasks and consolidated the
task-workspace resolver, and #1294 populated workspace_id in the claim
response for run_only autopilot tasks. Those closed the known fallthroughs
but the failure mode is still broader: whenever the daemon or server fails
to supply a workspace, the CLI silently falls back to
`~/.multica/config.json`, which is user-global, not workspace-scoped. On a
host running daemons for multiple workspaces, a single gap in workspace
propagation is enough to leak writes across workspaces.
This PR adds three coordinated guards so no single layer's bug can cause a
cross-workspace write:
1. `server/cmd/multica/cmd_agent.go` — `resolveWorkspaceID` detects the
agent execution context (`MULTICA_AGENT_ID` / `MULTICA_TASK_ID` env,
both daemon-only markers) and in that context refuses to fall back to
the user-global CLI config. Human / script usage (no agent env) is
unchanged: flag → env → config fallback chain still applies.
2. `server/internal/handler/daemon.go` — `ClaimTaskByRuntime` now
captures the runtime's workspace from `requireDaemonRuntimeAccess` and
enforces `resolved_task_workspace == runtime_workspace` after the
existing issue/chat/autopilot branches. On mismatch or empty, the
handler explicitly cancels the just-dispatched task (via
`TaskService.CancelTask`, which also reconciles agent status) and
returns 500. Without the explicit cancel, `ClaimTaskForRuntime` had
already transitioned the task to 'dispatched' and the agent status to
'working', so a plain 500 would leave both stuck for the ~5 min
stale-task sweep window.
3. `server/internal/daemon/daemon.go` — `runTask` refuses to spawn the
agent when `task.WorkspaceID` is empty (defense-in-depth against
server bugs and reused workdirs).
Tests:
- `cmd/multica/cmd_agent_test.go`:
`TestResolveWorkspaceID_AgentContextSkipsConfig` — five subtests
covering the full fallback matrix (outside agent context still reads
config; agent context uses env; agent context with empty env returns
empty; task-id-only marker also counts; requireWorkspaceID surfaces the
agent-context error message).
- `internal/handler/daemon_test.go`:
`TestClaimTaskByRuntime_TaskWorkspaceMismatch_CancelsAndRejects` —
constructs a data-inconsistent task (runtime_id in workspace A,
issue_id in workspace B) and asserts the handler returns 500 AND
leaves the task in 'cancelled' state (not 'dispatched').
Phase 1/2 follow-ups (prompt injection of workspace slug, session lookup
workspace filter, cross-workspace audit of agent-facing endpoints,
observability) are out of scope for this PR and tracked separately.
When both AWS_ENDPOINT_URL and CLOUDFRONT_DOMAIN are configured, the
uploaded file URL returned by S3Storage.Upload now uses the CDN domain
instead of the raw S3-compatible endpoint.
This enables S3-compatible backends (MinIO, R2, B2, Wasabi, etc.) to be
paired with a separate public-read domain — previously the CDN domain was
silently ignored whenever a custom endpoint was set, forcing clients to
hit the raw S3 API endpoint which typically requires signed requests.
No behavior change for deployments that set only one of the two vars:
pure AWS S3 with CloudFront, AWS S3 without a CDN, and MinIO/R2 without
a CDN all continue to return the same URLs as before.
The ListIssues and ListOpenIssues SQL queries omitted the description
column, so the API response never included description data. Board cards
checked issue.description (always null) and never rendered it, even when
the Description card property was enabled.
Add description to both SQL queries, the generated Go structs/scan calls,
and the response mapping functions.
The bluemonday HTML sanitizer applied to comment content (added in #679)
treats Markdown source as HTML, entity-encoding syntactically meaningful
characters and normalizing whitespace. This corrupts user input:
- "> quote" -> "> quote" (blockquote lost, see #1303)
- '"foo"' -> '"foo"' (literal entities visible)
- "\n\n2." -> " 2." (ordered list items merged into prose)
Comment content is stored as Markdown source. XSS is already handled at
two layers:
- Render: rehype-sanitize in packages/ui/markdown and
packages/views/editor/readonly-content (mention:// allowlist,
data-href restricted to http(s), class restricted to
code/div/span/pre).
- Edit: @tiptap/markdown is configured with html:false, so Markdown
source containing raw HTML tags is treated as plain text.
Removing the server-side sanitizer therefore does not lower the security
boundary, and restores faithful Markdown round-tripping.
The PR #1342 workaround in the editor serializer can be dropped once
this lands.
Co-authored-by: Eve <eve@multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(agent): add Kimi CLI as agent runtime
Adds support for Moonshot AI's Kimi Code CLI (https://github.com/MoonshotAI/kimi-cli)
as a new agent runtime, alongside Claude, Codex, OpenCode, OpenClaw, Hermes,
Gemini, Pi, Cursor and Copilot.
Kimi Code CLI implements the standard Agent Client Protocol (ACP) via the
`kimi acp` subcommand, so the new `kimiBackend` reuses the existing
hermesClient JSON-RPC transport in the agent package — only the binary,
client identity, log prefix, and tool-name extraction differ.
Wiring:
- server/pkg/agent: new kimiBackend + kimi_test.go; registered in New(),
LaunchHeader map, and the supported-types coverage test.
- server/internal/daemon/config.go: probes `kimi` (overridable via
MULTICA_KIMI_PATH / MULTICA_KIMI_MODEL).
- server/internal/daemon/execenv: writes AGENTS.md as the runtime context
file (Kimi reads AGENTS.md natively via /init), and writes skills under
`.kimi/skills/` so they are auto-discovered by the project-level skill
loader.
- packages/views/runtimes: ProviderLogo gains a Kimi mark.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* feat(agent/kimi): support per-agent model selection via ACP set_model
Wire Kimi into the model dropdown introduced in #1399:
- ListModels gets a 'kimi' case that drives the same ACP
initialize + session/new handshake as Hermes; both share a new
discoverACPModels helper and parseACPSessionNewModels parser
so future ACP backends only need a small provider entry.
- kimiBackend now issues session/set_model after session/new when
opts.Model is non-empty, mirroring the Hermes flow. Failures
fail the task instead of silently falling back to Kimi's
default model — silent fallback would hide that the dropdown
pick wasn't honoured.
Verified: go build ./..., go test ./pkg/agent/... ./internal/daemon/... ./internal/handler/..., pnpm typecheck and pnpm test (138 passed).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* refactor(agent): address code review feedback on Kimi runtime
- Share ACP provider-error sniffer between hermes and kimi. Previously
only hermes promoted stderr-observed 4xx/5xx into a failed task;
kimi would report "completed + empty output" when the Moonshot
upstream rejected a request (expired token, rate limit, …). Rename
hermesProviderErrorSniffer → acpProviderErrorSniffer and parameterise
the provider name; wire it into kimiBackend.Execute the same way.
- Rename extractHermesSessionID → extractACPSessionID (shared by all
ACP backends) so the name matches parseACPSessionNewModels.
- Drop the redundant second argument to kimiToolNameFromTitle; the
Message struct has only one relevant field (Tool), so passing it
twice was a dead fallback. Document that the function normalises
residual capitalised kimi titles not caught by hermesToolNameFromTitle.
- Remove kimi-only cmd.WaitDelay override; the hermes baseline is
fine for both and divergence adds noise.
- Add TestKimiBackendSetModelFailureFailsTask: fake `kimi acp` binary
that returns a JSON-RPC error for session/set_model, asserts that
the task result surfaces status=failed with the model name + upstream
message and preserves the session id.
- Fix stale agent listings in agent.go / daemon/config.go doc comments
(missing cursor, gemini, copilot).
All: `go build ./...`, `go vet ./...`, `go test ./pkg/agent/...
./internal/daemon/... ./internal/handler/...` green.
* fix(agent/kimi): pass --yolo so Shell tools don't hang on approval
Kimi's default config has `default_yolo = false`. Every Shell/file-mutating
tool call causes kimi acp to send a `session/request_permission` request
and block (up to 300s) waiting for a response. The daemon's hermesClient
only handles `session/update` notifications — permission requests go
unanswered, the tool call times out, and the UI loop eventually dies
("UI loop timed out"). Observed with the first real kimi task: agent sat
as Live for ~7 minutes before the daemon killed it.
The fix mirrors hermes' HERMES_YOLO_MODE=1 override: pass `--yolo` to
`kimi` so it auto-approves everything. `--yolo` is a top-level flag on
the `kimi` CLI (not a flag on `kimi acp`), so it must come before the
`acp` subcommand in argv. Added to kimiBlockedArgs so user custom_args
can't strip it.
While here, fix a related bug that made kimi tool names show up empty
in the daemon log ("tool #1: "): hermesToolNameFromTitle's fallback
returned `kind` when neither title-with-colon nor kind matched a known
tool. Kimi's ACP `tool_call` emits bare titles like "Shell" or "Read
file" with no `kind` at all, so we'd drop the title on the floor before
kimiToolNameFromTitle ever got a chance to map it. Now: preserve the
title when kind is unclassified; hermes titles always carry a colon so
this branch never fires for hermes.
Tests:
- TestKimiBackendPassesYoloFlag — fake binary that records its argv,
asserts --yolo comes before acp.
- TestHermesToolNameFromTitle rows for bare kimi-style titles.
- Existing suite green: go build, go vet, full pkg/agent + daemon +
handler test packages.
* fix(agent/acp): auto-approve session/request_permission from agent
The previous attempt (`kimi --yolo acp`) was a no-op. Inspected the
kimi-cli source: the `acp` Typer subcommand takes no parameters, so
flags on the root `kimi` command are dropped before `acp_main()` runs
— it's impossible to opt into YOLO mode through CLI flags for ACP.
The real fix is on our side: respond to session/request_permission.
ACP is bidirectional. When kimi runs a Shell or file-write tool, it
sends `session/request_permission` (agent → client, JSON-RPC request
with id + method) and waits up to 300s for a response. Our existing
hermesClient.handleLine only dispatched: (id + result/error) →
handleResponse, and (no id + method) → handleNotification. A request
with BOTH id and method fell through and got silently dropped — kimi
timed out, UI loop died, task sat stuck for 7 minutes.
Add handleAgentRequest: for session/request_permission, echo the id
and respond with outcome=selected, optionId=approve_for_session. The
daemon is headless; there's no user to prompt. `approve_for_session`
lets the agent remember the action so subsequent identical calls
(every Shell, every file write) skip the round-trip entirely. For any
other agent → client method, reply with standard -32601 method-not-
found so the agent doesn't block.
Also:
- Add writeMu so request() (main goroutine) and handleAgentRequest
(reader goroutine) don't interleave JSON frames on stdin.
- Revert the `--yolo acp` flag — it's a no-op, and carrying it in
kimiBlockedArgs gives the wrong impression that it does something.
Comment in kimi.go now points at handleAgentRequest as the real fix.
Tests:
- TestHermesClientAutoApprovesPermissionRequest: inject a
session/request_permission, assert the reply echoes the id and
carries {outcome: selected, optionId: approve_for_session}.
- TestHermesClientReplesMethodNotFoundForUnknownAgentRequest: confirm
unknown agent → client methods get JSON-RPC -32601 instead of silence.
- TestKimiBackendInvokesACPSubcommand replaces the yolo-flag assertion
with a negative assertion: no dead --yolo / --auto-approve / -y on
argv, since they'd pretend to do something they can't.
All: go build ./..., go vet ./..., go test ./pkg/agent/... green.
* fix(agent/acp): surface kimi tool input/output via content blocks
Kimi-cli emits tool_call and tool_call_update ACP frames with the
input/output inside a `content` array of ContentToolCallContent
blocks (shape: {type:"content", content:{type:"text", text:"..."}}),
not in the hermes-style `rawInput` map / `rawOutput` string. Our
parser only looked at rawInput/rawOutput, so the daemon recorded
empty Input and Output for every kimi tool — the execution-history
UI showed blank terminal panels even for commands that ran fine.
Add extractACPToolCallText() and a fallback in handleToolCallStart /
handleToolCallUpdate: when rawInput is nil / rawOutput is empty, pull
the text out of the content blocks. rawInput / rawOutput still take
precedence so hermes' behaviour is untouched. Terminal /
FileEditToolCallContent blocks are skipped (we have nothing to render
them as — kimi only emits TerminalToolCallContent when the client
advertises terminal capability, which we don't).
Tests:
- TestHermesClientHandleToolCallStartKimiContent — content array →
Input.text populated.
- TestHermesClientHandleToolCallCompleteKimiContent — multi-block
content → Output concatenated with newline separator.
- TestHermesClientHandleToolCallRawOutputTakesPrecedence — hermes
rawOutput still wins when both are present.
- TestExtractACPToolCallText — unit coverage for the helper
(single/multiple text blocks, terminal-block skip, empty input).
* fix(agent/acp): buffer streaming tool args so Input isn't empty in UI
kimi-cli streams tool args token-by-token via tool_call_update frames
— the initial tool_call carries an empty content block and each
subsequent in_progress update carries the cumulative JSON so far
(`{`, `{"comma`, `{"command": "echo`, …). The final completed update
then carries the tool's stdout, not the args. Observed per kimi-cli
acp/session.py::_send_tool_call{,_part,_result} and confirmed by
driving a real Shell call end-to-end: 10 in_progress frames, last
with `{"command": "echo hello world"}`, then completed with `hello
world\n`.
Our previous handleToolCallStart emitted MessageToolUse on the first
tool_call frame, capturing the empty content — so every kimi tool
appeared in the execution-history UI with a blank input. Output was
correct (fix 4335c198) but command was missing.
Changes:
- hermesClient now tracks pending tool calls per toolCallId. Hermes
path is unchanged — rawInput is present at tool_call time, so
emit-immediately-then-flag-emitted still fires on the initial frame.
- kimi path defers MessageToolUse until status=completed / failed.
tool_call_update in_progress frames update the buffered argsText
(cumulative, so overwrite); on completion we parse the accumulated
JSON into Message.Input. Malformed JSON falls back to `{"text": …}`
so non-JSON tool args still render.
- Orphan completion frames (no matching tool_call seen — e.g. daemon
restarted mid-task) synthesise ToolUse from the update's own
title/kind/rawInput so the UI still gets a header.
- extractACPToolCallText now also renders FileEditToolCallContent
blocks as a compact header ("--- path / +++ path / (edited: N → M
bytes)"). kimi emits these for Write / StrReplaceFile / Patch when
the tool's display block is a DiffDisplayBlock.
Tests:
- TestHermesClientKimiStreamingToolCall: empty tool_call + 5 streaming
in_progress + completed. Asserts no emission until complete, then
[ToolUse(Input.command="echo hi"), ToolResult(Output="hi\n")].
- TestHermesClientKimiMalformedArgsFallback: non-JSON argsText → falls
back to Input.text.
- TestHermesClientHandleToolCallCompleteOrphan: completed frame
without a start → ToolUse synthesised from update's rawInput.
- TestExtractACPToolCallText: diff + new-file-diff cases.
All agent / daemon / handler test packages green.
---------
Co-authored-by: Eve <8b0578a3-cf72-4394-9e38-b328eca92463@users.noreply.multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Eve <eve@multica.ai>
Co-authored-by: Lambda <f252c2c5-7d1d-4f3c-b394-a61abfe673fc@users.noreply.multica.ai>
Adds a first-class `model` field on agents so users can pick the LLM model from the create / settings UI instead of editing `custom_env` / `custom_args`. Each provider's dropdown is populated from the live CLI when possible (`opencode models`, `pi --list-models`, `openclaw agents list --json`, `cursor-agent --list-models`, hermes ACP `session/new` → `SessionModelState`), with a static catalog for providers that don't enumerate.
Daemon resolves the runtime model as `agent.model → MULTICA_<PROVIDER>_MODEL → ""` — empty passes through so each backend's CLI picks its own default, avoiding static-guess drift.
Per-provider honouring:
- Claude / Codex / OpenCode / Cursor / Gemini / Pi / Copilot — CLI `--model` / thread payload.
- OpenClaw — `opts.Model` is mapped to `--agent <name>` (the CLI rejects `--model`).
- Hermes — `session/set_model` ACP RPC; stderr is sniffed for provider-level errors so HTTP 4xx from the configured LLM surfaces instead of "empty output"; explicit-model failures mark the task `failed`.
Supporting changes: migration 050 adds `agent.model`; daemon ↔ server heartbeat piggyback carries a model-discovery request; new REST endpoints under `/api/runtimes/{id}/models`; `multica agent create --model` / `update --model`; shared `ModelDropdown` in `packages/views/agents` (searchable, creatable, provider-grouped, default-badge, runtime-supported gate).
The session cookie's Secure flag was tied to APP_ENV, and the
docker-compose self-host stack defaults APP_ENV to "production". On
plain-HTTP self-host deployments (LAN IP, private network) the browser
silently drops Secure cookies, leaving every subsequent /api/* call
anonymous and surfacing as 401 "auth: no token found" right after a
successful login.
Derive Secure from the scheme of FRONTEND_ORIGIN so HTTPS origins get
Secure cookies and plain-HTTP origins get non-secure cookies the
browser will actually store. Also harden cookieDomain() against the
other common trap: COOKIE_DOMAIN=<ip>, which RFC 6265 forbids and
browsers reject. Log a one-shot warning and fall back to host-only.
Docs: correct the COOKIE_DOMAIN description (it was labelled as
CloudFront-only but applies to session cookies too) and call out the
IP-literal pitfall in SELF_HOSTING_ADVANCED.md, self-hosting.mdx, and
.env.example.
Refs #1321
AgentLiveCard kept its taskStates map across issueId prop changes, and
its merge logic only added newly-fetched tasks without removing stale
ones. Navigating from Issue A (with a running agent) to Issue B via
cmd+k left A's sticky agent status card pinned on B's page.
Key AgentLiveCard and TaskRunHistory by issue id so React remounts
them when the issue changes, guaranteeing fresh state per issue.
Closes MUL-1147
Agents were silently finishing tasks without ever posting results to the
issue — their final reply stayed in terminal/log output only. See MUL-1124.
Root cause: the injected CLAUDE.md / AGENTS.md put "post a comment with
results" inside the body of step 4 (a nested clause in the default workflow
description), so skill-driven flows jumped straight from "do the work" to
`status in_review`.
- Hoist posting the result comment into its own explicit, numbered step in
both assignment-triggered and comment-triggered workflows, with the exact
`multica issue comment add` invocation inlined.
- Add a hard warning at the top of the Output section that terminal / chat
text is never delivered to the user.
- Add regression test covering both workflow branches.
* feat(server): configurable pgxpool size with sane defaults
pgxpool.New(ctx, url) silently sets MaxConns = max(4, NumCPU). On the
prod pods that resolved to 4, which got fully saturated by daemon
claim/heartbeat traffic (~3800 acquires/s) and showed up as ~900ms
acquire waits on every query — the actual root cause of the 3s+
/tasks/claim tail latency. The db pool stats logging from #1378
confirmed this with empty_acquire_delta == acquire_count_delta.
Switch to pgxpool.ParseConfig + NewWithConfig and apply per-pod
defaults of MaxConns=25 / MinConns=5, both overridable via env vars
(DATABASE_MAX_CONNS / DATABASE_MIN_CONNS) so the size can be tuned
in prod without a redeploy.
The defaults follow the standard 'small pool, lots of waiters' guidance
for Postgres (PG community / HikariCP formula
`(core_count * 2) + effective_spindle_count`); 25 leaves headroom for
bursts and occasional long queries while staying safely under typical
managed-Postgres max_connections ceilings when multiplied across pods.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(server): respect DATABASE_URL pool_* params; add precedence tests
Address review feedback on #1381:
- Configuration precedence is now explicit: DATABASE_MAX_CONNS env >
pool_max_conns query param on DATABASE_URL > built-in default. Same
for min_conns. Previously the env-empty path unconditionally
overwrote whatever ParseConfig had read from the URL — a silent
regression for deployments that already tuned pool size via the
connection string.
- Add unit tests in dbstats_test.go covering each precedence branch
(defaults, URL-only, env-over-URL, partial URL, invalid env,
min>max clamp).
- Move pool tuning vars out of 'Required Variables' into a new
'Database Pool Tuning (Optional)' section in SELF_HOSTING_ADVANCED.md
so self-hosters don't think they need to set them.
- Add commented entries in .env.example.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(server): invalid pool env falls back to URL/code default, never pgx 4
Address second round of review on #1381:
Previous code passed cfg.MaxConns / cfg.MinConns as the envInt32 fallback,
which meant an invalid DATABASE_MAX_CONNS value silently fell back to
ParseConfig's value — i.e. pgx's built-in default of 4/0 when the URL had
no pool_* params. That's exactly the bad value this PR exists to remove,
and the previous test (TestPoolSizing_InvalidEnvFallsBack) accidentally
locked it in.
Compute the non-env fallback first (URL pool_* if present, else code
default 25/5) and pass that to envInt32. Misconfigured env now lands on
the same value as if the env were unset — never on the pgx default.
Replace the loose 'max > 0' assertion with two precise tests:
- invalid env + no URL param → code default (25/5)
- invalid env + URL param → URL value
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
After merging the per-phase claim slow-logs (#1376), the prod data showed
the smoking gun: unrelated endpoints (claim, heartbeat, /api/workspaces,
ping) all completed at the *same wall-clock instant* with durations
clustered at ~1.4s and ~2.88s — and within the claim breakdown,
list_pending_ms was 713ms even when list_pending_count=0.
A 0-row indexed scan can't take 713ms, and unrelated endpoints don't
synchronize their completion by accident. The only explanation that fits
is requests blocking on a shared resource and being released together.
The most likely culprit is pgxpool connection-acquire wait: pgxpool.New
is called with no config, so MaxConns defaults to max(4, NumCPU) — under
the daemon poll fan-in this is trivially exhausted.
This change adds the observability needed to confirm/refute that without
changing pool sizing yet (pool sizing is a follow-up once we have data):
- logPoolConfig: prints MaxConns / MinConns / MaxConnLifetime /
MaxConnIdleTime / HealthCheckPeriod once at startup. Surfacing the
effective limit is critical because the default is surprisingly small
and easy to mistake for 'database is slow'.
- runDBStatsLogger: samples pool.Stat() every 15s (matches daemon
heartbeat cadence for easy correlation). Emits INFO with TotalConns /
AcquiredConns / IdleConns and per-window deltas (acquire_count,
empty_acquire, canceled_acquire, avg_acquire_ms). Auto-upgrades to
WARN whenever empty_acquire_delta > 0 or canceled_acquire_delta > 0
— those are the direct symptom of a request having to wait because
no idle connection was available.
If on prod we see 'db pool pressure' WARN lines coincident with the
claim_endpoint slow lines, the hypothesis is confirmed and the fix
becomes a one-liner (pool config tuning + the existing N+1 reduction
ideas to lower demand).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* chore(server): add slow-path timing logs for /tasks/claim
We're seeing 3s+ tail latency on POST /api/daemon/runtimes/{rid}/tasks/claim
in production. Before changing the code, add structured timing logs along
the entire claim path so we can confirm where the time is actually going.
Three layers, all gated by a slow-only threshold to avoid log spam at the
default 3s daemon poll cadence:
- handler.ClaimTaskByRuntime (>=500ms): splits auth_ms / claim_ms /
build_ms so we can tell whether the slowness is in the actual claim
query or the post-claim response assembly (GetAgent, LoadAgentSkills,
GetIssue, GetWorkspace, GetComment, GetLastTaskSession, or the chat
branch's 4 queries).
- service.ClaimTaskForRuntime (>=300ms): logs list_pending_ms,
list_pending_count, agents_tried, claim_loop_ms — directly validates
the suspected N+1 amplification (one ListPendingTasksByRuntime + one
ClaimTask per unique agent).
- service.ClaimTask (>=300ms): splits get_agent_ms / count_running_ms /
claim_agent_ms so we can isolate the NOT EXISTS + FOR UPDATE SKIP
LOCKED cost from the surrounding metadata reads.
Pure observability change. No behavior change in the request path.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* chore(server): widen claim slow-log to cover post-claim DB work and error paths
Address review feedback on #1376: the previous version emitted
'claim_task slow' before updateAgentStatus and broadcastTaskDispatch,
both of which can hit the DB (broadcastTaskDispatch goes through
ResolveTaskWorkspaceID and may re-query issue/chat_session/autopilot_run).
That meant a claim that was actually slow in the post-claim tail would
either be under-counted or not logged at all, defeating the purpose of
the instrumentation.
Changes:
- ClaimTask: switch to defer-based exit logging. Adds update_status_ms
and dispatch_ms phase fields. Error paths now also emit a slow log
with outcome=error_get_agent / error_count_running / error_claim.
- ClaimTaskForRuntime: same defer pattern; error paths log with
outcome=error_list / error_claim, partial loop time still captured.
- ClaimTaskByRuntime handler: same defer pattern; auth-failure / claim-
error paths now also carry phase timings (outcome=unauth / error_claim).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(agent/comments): re-emit trigger comment id every turn + server-side parent_id guard
Resumed Claude sessions keep prior turns' tool calls in context, so a
comment-triggered task could reuse the PREVIOUS turn's --parent UUID
instead of the current trigger's. The reply landed in the wrong thread
(MUL-1125): backend stored exactly what the agent sent, but the agent
pulled a stale UUID from its own conversation memory.
Two layers of defense:
1. Extract BuildCommentReplyInstructions so daemon.buildCommentPrompt
and execenv.InjectRuntimeConfig emit the same "use this exact
--parent, do not reuse values from previous turns" block. The
per-turn prompt now carries the current TriggerCommentID, which it
previously relied on CLAUDE.md for (and CLAUDE.md isn't re-read
mid-session).
2. Handler-side guard in CreateComment: when an agent posts from inside
a comment-triggered task (X-Agent-ID + X-Task-ID, task has
TriggerCommentID), require parent_id == task.TriggerCommentID or
return 409. Assignment-triggered tasks are untouched.
* fix(agent/comments): scope parent_id guard to the task's own issue
Two issues from CI + GPT-Boy's review:
1. Guard was too broad: the CLI stamps X-Task-ID on every request, so an
agent legitimately commenting on a different issue while its current
task was comment-triggered would get 409'd with the wrong issue's
trigger comment id. Narrow the guard to fire only when the request's
issue matches the task's own issue — cross-issue agent activity
stays unblocked.
2. The integration test tried to insert a second queued task for the
same (agent, issue), which hits the idx_one_pending_task_per_issue_agent
unique index. Replace the assignment-triggered-task sub-case with a
cross-issue regression test (the scenario we now need to cover anyway):
post on issue B while X-Task-ID points at a comment-triggered task on
issue A, expect 201.
Replace the single day picker in the "Weekly" frequency with a multi-select
so users can schedule on any combination of weekdays (e.g. Mon/Wed/Fri)
in addition to the existing "Weekdays" Mon-Fri preset.
The backend already accepts any day-of-week list in the cron expression,
so this is a frontend-only change. Relabels the tab to "Days" to reflect
the new behavior.
* fix(agent/codex): surface stderr tail in initialize / turn startup errors
When codex app-server exits before the JSON-RPC handshake completes —
e.g. because the user put a flag in custom_args that the subcommand
rejects — the Result.Error users see is `codex initialize failed:
codex process exited`, while codex's actual complaint (typically
something like `error: unexpected argument '-m' found`) only lives in
daemon logs.
Wrap the stderr writer with a bounded stderrTail that still forwards
to the slog logWriter but also retains the last 2 KiB of bytes
written. Include that tail on the three startup failure paths
(initialize, startOrResumeThread, turn/start). Runtime cancellation
paths are left untouched — they're our own abort and the stderr
context isn't a clear signal there.
Refs #1308. Complement to #1310 / #1312 — lets "bad custom_args fail
loudly" actually be workable by giving the failure a real message.
* fix(agent/codex): join cmd.Wait() before sampling stderr tail
Addressing review of #1314: reading stderrBuf.Tail() right after
c.request returns "codex process exited" was racy. Nothing in that
path synchronizes with os/exec's internal stderr copy goroutine —
cmd.Wait() is the only documented join point. The original defer ran
cmd.Wait() later, but by then we had already built Result.Error from
a potentially-empty Tail().
Replace the ad-hoc deferred stdin.Close()/cmd.Wait() with a
sync.Once-wrapped drainAndWait closure. Call it explicitly on the
three startup failure paths before sampling the tail; keep it as the
cleanup defer so the success path behaves identically.
Also add TestCodexExecuteSurfacesStderrWhenChildExitsEarly: spawns a
real subprocess that prints to stderr and exits before responding to
initialize, runs it through Execute, and asserts Result.Error
contains the stderr hint. This covers the full timing path the
reviewer flagged, which the helper-level tests in this PR did not.
* feat(desktop): hourly update poll + manual check button in settings
The previous updater only ran one check 5s after launch, so a missed
or failed initial check meant the user had to fully restart the app to
see a new release. Add a 1h background poll for long sessions and a
"Check now" button under a new Updates tab in Settings so the user can
trigger a check on demand without waiting.
The button reuses the existing autoUpdater pipeline — when an update is
available the existing corner notification still drives the download
flow; the settings tab only surfaces the immediate check result
(up-to-date / available / error).
* fix(desktop): trust electron-updater's isUpdateAvailable for the manual check
Per review: deriving `available` from a version-string compare is wrong —
`updateInfo.version` can differ from `app.getVersion()` while
electron-updater still suppresses `update-available` (pre-release channels,
staged rollouts, downgrade scenarios, min-system-version gates). In those
cases the settings tab would say "vX is available" but no corner download
prompt would ever appear. Use `result?.isUpdateAvailable` instead, which is
electron-updater's own answer.
Follow-up to PR #1188 / migration 047, which intentionally omitted the
five historical conflict slugs (admin / multica / new / setup / www) from
the reserved-slug audit because each had one production workspace using
it at the time and we did not want to block deploy on owner outreach.
MUL-972 closed that loop on prd for four of the five:
* admin (99cd10e4-…) → renamed to legacy-admin-99cd10e4
* multica (dcd796aa-…) → renamed to legacy-multica-dcd796aa
* new (e391e3ed-…) → renamed to legacy-new-e391e3ed
* www (5e8d38b2-…) → workspace deleted (was empty: 0 issues /
projects / agents, owner-only member; 18
workspace-FK relations all CASCADE)
This PR:
1. Adds migration 049_audit_legacy_reserved_slugs which audits those
four slugs against workspace.slug at startup. If any future workspace
slips in with one of them, startup fails loudly via RAISE EXCEPTION
instead of being silently shadowed by a global route. Mirrors the
structure of 047.
2. Adds 'multica' / 'www' / 'new' to the reserved-slug allow-deny list
in both the Go handler and the shared TS list (admin was already in
both). Keeps the two lists in lockstep per the convention enforced
in workspace_reserved_slugs.go header.
setup is STILL exempt from the audit and is intentionally NOT added to
the reserved list. The setup workspace (b43f0bc2-…) is a real production
user (owner: Roberto Betancourth, building a chants/Alabanzas app) and
is being handled out-of-band via owner outreach. A separate follow-up
migration will fold setup into the audit once that workspace's slug has
been migrated.
Migration is intentionally shipped AFTER the prd data fix (not before):
049 will RAISE EXCEPTION on any remaining conflict, so we want the data
state clean first. Rollout order:
prd data fix (done by db-boy on 2026-04-20) → this PR.
Tested:
- go test ./server/internal/handler/ -run TestReserved → pass
- pnpm --filter @multica/core test consistency → pass (4/4 in
consistency.test.ts; global-prefix↔reserved invariant holds)
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
When Desktop opens /login?platform=desktop in the browser and the user
already has a valid web session, the page previously bounced them to
their workspace and Desktop never received a token. Now we mint a bearer
token via issueCliToken and redirect through the multica:// deep link so
Desktop completes sign-in without a second Google round-trip.
Refs: MUL-1080
Fixes#1332.
Two regressions introduced in #910 (2026-04-14, "OpenClaw backend P0+P1
improvements") that together block all openclaw users:
1. `openclaw agent` does not accept `--model` or `--system-prompt`, so
any agent configured with a Model field crashed in ~700ms with
`exit status 1`. Remove both forwards, and add them to
openclawBlockedArgs so custom_args can't reintroduce the crash.
Model is bound at registration time via `openclaw agents
add/update --model`.
2. AgentInstructions were written to `{workDir}/AGENTS.md` by
execenv.InjectRuntimeConfig, but openclaw loads bootstrap files
from its own workspace dir — the file was never read, so every
agent's Instructions field was silently discarded. Populate
opts.SystemPrompt for the openclaw provider in runTask and
prepend it to the `--message` payload in the backend so the
model actually receives the instructions.
Other providers surface instructions through their native runtime
config file (CLAUDE.md / AGENTS.md / GEMINI.md) and are intentionally
left unchanged to avoid double injection.
Extract buildOpenclawArgs so arg construction is directly testable;
add unit tests covering the removed flags, the SystemPrompt prepend,
and custom_args filtering.
* fix(chat): preserve chat session resume pointer across failures
The chat 'forgets earlier messages' bug came from PriorSessionID being
silently lost in several edge cases:
- UpdateChatSessionSession unconditionally overwrote chat_session.session_id,
so any task that completed without a session_id (early agent crash,
missing result) wiped the resume pointer to NULL.
- CompleteAgentTask + UpdateChatSessionSession ran in separate calls. A
follow-up chat message claimed in between resumed against a stale (or
NULL) session and started over.
- FailAgentTask never wrote session_id back, so a task that established
a real session before failing lost its resume pointer.
- ClaimTaskByRuntime only trusted chat_session.session_id and never
fell back to the existing GetLastChatTaskSession query, so a single
bad turn could permanently drop the conversation memory.
This change:
- Use COALESCE in UpdateChatSessionSession so empty inputs preserve the
existing pointer; surface DB errors instead of swallowing them.
- Run CompleteAgentTask/FailAgentTask + UpdateChatSessionSession inside
the same transaction (TaskService now takes a TxStarter).
- Extend FailAgentTask + the daemon FailTask path (client, handler,
service) to forward session_id/work_dir, so failed/blocked tasks that
built a real session still record it.
- Fall back to GetLastChatTaskSession in ClaimTaskByRuntime when the
chat_session pointer is missing, and include failed tasks in that
lookup so a single failure can't lose the conversation.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(daemon): forward session_id/work_dir on blocked + timeout paths
runTask previously dropped result.SessionID and env.WorkDir on the
non-completed return paths:
- timeout returned a naked error, so handleTask called FailTask with
empty session info and the chat resume pointer was either left stale
or eventually overwritten with NULL.
- blocked / failed (default branch) returned a TaskResult without
SessionID / WorkDir, so even though FailTask now COALESCEs into
chat_session, there was no value to write through.
- the empty-output completion path was the same: it raised an error
even when a real session_id had been built.
All three paths now return a TaskResult that carries the SessionID /
WorkDir the backend produced. Combined with the COALESCE-based update
in UpdateChatSessionSession and the FailTask plumbing introduced in
PR #1360, the next chat turn can always resume from the latest agent
session — even when the previous turn timed out, was rate-limited, or
returned an empty completion — instead of starting over with no memory
of the conversation.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(copilot): capture session id from session.start as fallback
The Copilot backend only read sessionId from the synthetic 'result'
event, ignoring the one already present on session.start. When the CLI
was killed before result arrived (timeout, cancel, crash, or a
session.error mid-turn), the daemon reported SessionID="" and the
chat-session resume pointer could not advance — causing the chat to
silently drop conversation memory on the next turn.
Capture session.start.sessionId into state up front, and only let
'result' overwrite it when it actually carries one. result still wins
when present (it is the authoritative end-of-turn record).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(copilot): parse premiumRequests as float to preserve session id
Copilot CLI v1.0.32 serializes premiumRequests as a float (e.g. 7.5),
not an integer. Our copilotResultUsage struct typed it as int, which
made the entire 'result' line fail json.Unmarshal — silently dropping
sessionId on every turn.
This was the real cause of chat memory loss: the daemon reported
SessionID="" to the server, chat_session.session_id stayed NULL, and
the next chat turn never received --resume <id>, so each turn started
a fresh Copilot session with no prior context.
Add a regression test using the real JSON line from CLI v1.0.32 that
asserts sessionId is preserved when premiumRequests is fractional.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Eve <eve@multica.ai>
Co-authored-by: yushen <ldnvnbl@gmail.com>
The v0.2.6 self-host security fix (#1307) defaults APP_ENV to "production"
in docker-compose.selfhost.yml, which disables the 888888 master verification
code. The follow-up docs PR (#1313) covered SELF_HOSTING.md and the
installers, but `.env.example` — the file users actually copy and edit —
still makes no mention of APP_ENV, so operators who don't read the prose
docs hit the exact same "888888 stopped working after upgrade" confusion
reported in #1331.
- Add APP_ENV= to .env.example with a comment block that spells out the three
cases (Docker default, local dev, evaluation) and warns against enabling
the bypass on public instances. Keeping the value empty preserves the
current `make dev` UX (Go server reads empty → treats as non-production →
888888 works locally) while `${APP_ENV:-production}` in the compose file
still ensures public Docker deployments are safe by default.
- Update the existing 888888 mention under # Email so it no longer
contradicts the new gating rule.
- Update the `make selfhost` post-start banner, which still told operators
to "Log in with any email + verification code: 888888" even after #1307
disabled that path by default.
Adds a "Create sub-issue from selection" button to the editor bubble
menu. When an issue context is present (description editor, comment
input, reply input, comment edit), selecting text and clicking the
button creates a new issue parented under the current issue and
replaces the selection with a mention link to the new issue.
Closes#930
- Added environment variables to control signups
- Updated frontend to hide signup text when disabled
- Added backend check to block new user creation via magic link
- Updated .env.example
ESC did nothing inside the issue description editor because browsers
don't blur contenteditable elements by default, leaving users stuck in
the editor with no keyboard escape hatch.
Add a blur-shortcut extension mirroring TitleEditor's behavior and wire
it into ContentEditor's edit-mode extension set.
Previously the issue detail top bar only showed 'workspace name > identifier'.
Add the issue title next to the identifier so users can see what issue they're
viewing without scrolling.
* fix(sidebar): stabilize useQuery default arrays to prevent render loop
Inline `= []` defaults on `useQuery` return a new array reference on
every render when `data` is undefined (query disabled or mid-load).
Downstream effects/memos that depend on the value then fire every
render; the pinned-items `useEffect` compounds this by calling
`setLocalPinned` each time, so under sustained `data === undefined`
(e.g. backend unreachable, WebSocket in reconnect loop) React trips
its "Maximum update depth exceeded" guard and the sidebar becomes
unusable.
Use module-level empty-array constants so the default identity stays
stable across renders.
* fix(chat): short-circuit ResizeObserver update when bounds unchanged
The resize observer always called `setRevision(r => r + 1)` from its
callback, even when `clientWidth`/`clientHeight` were identical to the
previous reading. Any spurious notification — sub-pixel layout jitter
during mount, or an ancestor reflow triggered by an unrelated state
update — then fed back into the same render cycle and could exceed
React's update-depth limit.
Guard the state bump by comparing against the previous bounds, and
leave `setBoundsReady(true)` outside the guard since it's idempotent.
Following #1307, the Docker self-host stack defaults to APP_ENV=production,
which disables the 888888 master verification code on auth.go:169. The
installer banners and self-hosting docs still told operators to log in with
888888, leaving them stuck.
Update install.sh, install.ps1, SELF_HOSTING.md, SELF_HOSTING_ADVANCED.md,
and self-hosting.mdx to document the three login paths: configure
RESEND_API_KEY (recommended), set APP_ENV=development to enable 888888 for
private evaluation, or read the dev verification code from backend container
logs. Also warn against enabling APP_ENV=development on public instances.
* fix(agent/codex): route custom_args -m/--model to thread/start payload
Codex agents spawn via `codex app-server --listen stdio://`, which does
not accept `-m` / `--model` (those belong to the normal Codex CLI). When
a user's custom_args still carried those tokens the process exited
before the JSON-RPC initialize handshake with `codex process exited`,
with no actionable error.
Extract `-m <v>`, `--model <v>`, and `--model=<v>` from opts.CustomArgs
before invoking app-server and promote the value into opts.Model, so
that startOrResumeThread can pass it through the `thread/start` payload
where Codex actually reads the field.
Fixes#1308.
* fix(ui/agents): drop Codex-incompatible --model example from custom args tab
The helper text and placeholder suggested `--model claude-sonnet-4-…` as
a custom CLI argument, which is valid for Claude but crashes Codex
agents (its `app-server` subcommand does not accept model flags). Swap
in provider-agnostic copy so the UI no longer steers users into an
invalid configuration for non-Claude runtimes.
Refs #1308.
* revert "fix(agent/codex): route custom_args -m/--model to thread/start payload"
This reverts f18355b2. After review, extracting `-m`/`--model` out of
opts.CustomArgs and promoting them into the thread/start payload is the
wrong shape of fix: agent CLIs have many flags their non-interactive
modes don't accept, and hand-translating a subset case-by-case doesn't
scale — it pushes us toward an ever-growing list of per-backend arg
rewriters.
The preferred direction is to teach users via the UI what command their
custom_args extend (see the launch_header preview in #1312) and let
bad configurations fail loudly. If the resulting error is hard to read
that's a separate improvement we should make on the failure path, not
by silently rewriting user input.
Refs #1308.
* feat(agent): add LaunchHeader per agent type
Each backend in server/pkg/agent/ hardcodes a stable command skeleton
(e.g. `codex app-server --listen stdio://`, `hermes acp`) before
appending opts.CustomArgs. Surfacing that skeleton lets the UI tell
users which command their custom_args are being appended to, so a
Codex user doesn't mistakenly add `-m gpt-5.4-mini` expecting it to
reach the CLI when the subcommand is actually `app-server`.
Expose only the minimum that aids judgment — binary + subcommand, or a
short mode label when there is no subcommand — and deliberately omit
transport values, internal flags, and env to keep the surface small
and renaming-safe.
Refs #1308.
* feat(handler/runtime): surface launch_header on runtime response
runtimeToResponse now derives launch_header from agent.LaunchHeader,
piggybacking on the runtime's existing provider field so the
frontend's RuntimeDevice gains the skeleton without a new endpoint or
DB query. Client gets the header for free whenever it lists agents'
runtimes — which the custom-args tab already does.
Refs #1308.
* feat(ui/agents): show launch mode preview in custom args tab
Thread the resolved RuntimeDevice from AgentDetail into CustomArgsTab
and render its launch_header as a one-line preview above the args
list, so users see `codex app-server <your args>` (or equivalent per
provider) and can tell whether a CLI-style flag like `--model` will
actually reach the invoked subcommand. Source of truth stays in the
Go backend; the TS type just carries the string.
Refs #1308.
* refactor(auth): add sanitizeNextUrl helper in @multica/core/auth
Extracts a reusable helper that returns a post-login redirect URL only
when it's a safe single-slash relative path, and null otherwise. Rejects
absolute URLs, protocol-relative URLs, backslashes, and control
characters so call sites can safely pass the result to router.push().
Keeping the rule in a single helper (with direct unit tests) avoids
each consumer re-implementing the validation and drifting.
* fix(auth): validate next= redirect target to prevent open redirect
Closes#1116
Next.js router.push accepts absolute URLs, so a crafted
`/login?next=https://evil.example` would send the user off-origin
after a successful login. The Google OAuth callback has the same
vector via the `state=next:<url>` payload.
Sanitize both entry points through `sanitizeNextUrl` from
`@multica/core/auth` so only safe single-slash relative paths survive;
null results fall through to the existing workspace-list-based default
without any hard-coded path.
---------
Co-authored-by: JunghwanNA <70629228+shaun0927@users.noreply.github.com>
* fix(comment): assignee on_comment path should use reply id, not thread root
Symmetric fix to #871 — that PR fixed the @mention path but missed the
assignee on_comment path in the same file. Replies on agent-assigned
issues were still getting trigger_comment_id = parent_id, so the daemon
fed the parent comment's content to the resumed claude session, which
then either exited with 'Already replied to comment <parent>' or silently
misrouted its answer depending on model / session state.
Reply placement (flat-thread grouping) is already decoupled from
trigger_comment_id by TaskService.createAgentComment's parent
normalization (added alongside #871), so passing comment.ID directly is
safe and matches the mention path's post-#871 behavior.
Fixes#1301
Made-with: Cursor
* test(comment): assert assignee on_comment records reply id as trigger_comment_id
Integration regression guard for #1301. Asserts that after a member posts
a reply under an agent-authored thread, the enqueued agent task's
trigger_comment_id matches the new reply, not the thread root. Without
the companion fix in comment.go the old parent-override would store the
root id and the daemon would feed stale content (via prompt.go
BuildPrompt) to the agent.
Made-with: Cursor
---------
Co-authored-by: fuxiao <fuxiao@zyql.com>
Agent mentions enqueue a new task; member mentions send a notification.
Without this warning, agents have used `[@Name](mention://agent/<id>)` in
prose (e.g. "GPT-Boy is correct") and accidentally re-triggered the agent.
Adds a caveat under `## Mentions` in the prompt injected into agent
runtimes, plus tightens the Agent bullet to make the side-effect explicit.
The autopilot detail page mapped `status: "running"` to a `Loader2` icon
but rendered it without `animate-spin`, so a manually-triggered run sat
on a static circle until the row flipped to completed/failed and the
user got no visual feedback that anything was happening.
Add an optional `spin: true` flag to the run-status config and apply
`animate-spin` when set. Only the running entry is marked.
When --resume targets a dead session, claude prints
"No conversation found with session ID: ..." to stderr, emits a stream-json
system init with a fresh session_id, then exits with code 1. The backend
was treating that fresh id as the authoritative session, so
daemon.go's retry-with-fresh-session fallback (SessionID == "" guard)
never triggered. Every subsequent task for the same (issue, agent) pair
stayed permanently broken until the server-side session_id was cleared by
hand.
Fix: when --resume was requested but the emitted session_id differs AND
the run failed, drop the fresh id from Result so the daemon's existing
fallback can do its job. Factored into a pure helper and unit-tested.
Fixes#1284
Co-authored-by: fuxiao <fuxiao@zyql.com>
* fix(agent): add per-agent mcp_config field to restore MCP access
Closes#1111
The --strict-mcp-config flag was added defensively in #592 to prevent
Claude agents from inheriting MCP state from the outer Claude Code session.
It was meant to be paired with --mcp-config <path> to inject a controlled
set of MCPs, but that path was never implemented, which silently stripped
all user-scope MCPs from spawned agents.
This PR completes the original design by:
- Adding a nullable mcp_config jsonb column to the agents table
- Wiring mcp_config through AgentResponse, Create/Update requests
- Piping it into ExecOptions.McpConfig in the daemon
- Serializing to a temp file and passing --mcp-config <path> in buildClaudeArgs
- Blocklisting --mcp-config in claudeBlockedArgs to prevent override
via custom_args
Does not touch Codex provider (tracked separately in #674).
Does not implement Multica MCP auto-injection (out of scope).
* fix: disambiguate JSON null vs absent for mcp_config
The release workflow previously triggered on 'v*', which matched a
stray 'v0.2.5-dirty' tag pushed to the repository. GoReleaser ran
again and overwrote the Homebrew formula with a 0.2.5-dirty version
whose tarball URLs 404.
Tighten the trigger to semver-shaped tags and add an explicit guard
that fails the job if the tag name contains '-dirty' (which can come
from 'git describe --tags --dirty').
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Docs site is no longer auto-deployed via Vercel (disabled in
dashboard), so building it on every PR adds friction without
catching anything actionable. Use turbo's negative filter to
skip @multica/docs across all three tasks.
Self-hosted services (postgres, backend, frontend) should restart
automatically on failure or host reboot. This is standard practice
for production docker-compose deployments.
Co-authored-by: Zhazha <zhazha@openclaw.internal>
- Mark AppLink draggable={false} and add pointer-events-none while
dragging, so the browser's native <a> drag (which otherwise navigates
to the pin's href on mouse release) is suppressed.
- Introduce a component-local pinnedItems snapshot gated by an
isDraggingRef, so a mid-drag TQ cache write (optimistic or WS
refetch) cannot reorder the DOM under dnd-kit's drop animation.
Mirrors the pattern already used by board-view.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
GitHub Copilot CLI scans project-level skills from .github/skills/<name>/SKILL.md
(per the official cli-config-dir-reference docs), not from .agent_context/skills/.
Previously, skills injected for the copilot provider were placed under
.agent_context/skills/ and only referenced by name in AGENTS.md, meaning
Copilot would not actually pick them up.
- resolveSkillsDir: add a dedicated copilot case writing to .github/skills/
- Update doc comments in context.go and runtime_config.go
- Add TestWriteContextFilesCopilotNativeSkills covering the new path and
ensuring .agent_context/skills/ is not created for copilot
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
electron-builder 26.8.1 rejects publishingType under the GitHub publisher;
the correct option for selecting draft/prerelease/release is releaseType.
Using publishingType caused schema validation to fail during packaging.
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(agents): make issue tasks easier to open from agent details
Make task rows in the Tasks tab navigate directly to the related issue
detail page when issue data is available, using AppLink for cross-platform
compatibility. Rows without resolved issue data remain non-clickable.
Adds a subtle hover shadow to make the interactive area more discoverable.
Closes#1129
* fix(agents): use workspace issue paths in tasks tab
* test(agents): cover tasks tab issue links
* feat(docs): mount docs site at /docs subpath via basePath + multi-zone
Configure the Fumadocs site so it can be served at multica.ai/docs:
- Add basePath: '/docs' to apps/docs/next.config.mjs
- Flatten routes: drop standalone home, render content/docs/index.mdx at
the root, move catch-all from app/docs/[[...slug]] to app/[...slug]
- Wrap children with DocsLayout in the root layout (was a separate
segment-level layout under app/docs/)
- Set source loader baseUrl to '/' so URL slugs no longer carry the
basePath (Next.js prepends it automatically)
- Strip the now-redundant '/docs/' prefix from internal MDX links and
drop the duplicate "Documentation" nav entry
- Add app/not-found.tsx for App Router 404 handling
Wire up multi-zone routing so apps/web proxies /docs/* to the docs app:
- Add DOCS_URL env (default http://localhost:4000) and rewrites for
/docs and /docs/:path* in apps/web/next.config.ts
- Whitelist DOCS_URL in turbo.json globalEnv
* fix(web): move /docs rewrite to beforeFiles so [workspaceSlug] doesn't shadow it
The /docs rewrite was running in the default afterFiles slot, which is
evaluated *after* file-system routing. apps/web/app/[workspaceSlug]/
matched /docs first as a workspace named "docs" (which doesn't exist) and
returned 404 before the rewrite to the docs Vercel project ever fired.
Splitting rewrites into beforeFiles/afterFiles puts /docs and
/docs/:path* ahead of route resolution so they always proxy to the docs
zone.
* feat(cli): add `issue subscriber` commands
Wrap the existing /subscribers, /subscribe, and /unsubscribe endpoints as
`multica issue subscriber list|add|remove`, mirroring the comment subcommand
shape. `--user <name>` reuses resolveAssignee to resolve a member or agent;
without the flag, the action targets the caller.
* fix(issues): default subscribe target to resolveActor, not X-User-ID
When no user_id is posted, subscribe/unsubscribe hardcoded the target as
("member", X-User-ID). A CLI caller running as an agent (X-Agent-ID set)
then subscribed the underlying member rather than the agent itself,
which contradicts the "defaults to the caller" contract.
Derive the default via resolveActor so the endpoint mirrors caller
identity consistently — agent caller → agent row, member caller →
member row. Adds a regression test covering the agent caller path.
Configure the Fumadocs site so it can be served at multica.ai/docs:
- Add basePath: '/docs' to apps/docs/next.config.mjs
- Flatten routes: drop standalone home, render content/docs/index.mdx at
the root, move catch-all from app/docs/[[...slug]] to app/[...slug]
- Wrap children with DocsLayout in the root layout (was a separate
segment-level layout under app/docs/)
- Set source loader baseUrl to '/' so URL slugs no longer carry the
basePath (Next.js prepends it automatically)
- Strip the now-redundant '/docs/' prefix from internal MDX links and
drop the duplicate "Documentation" nav entry
- Add app/not-found.tsx for App Router 404 handling
Wire up multi-zone routing so apps/web proxies /docs/* to the docs app:
- Add DOCS_URL env (default http://localhost:4000) and rewrites for
/docs and /docs/:path* in apps/web/next.config.ts
- Whitelist DOCS_URL in turbo.json globalEnv
Before this PR, `EnsureDaemonID(profile)` wrote to ~/.multica/profiles/
<profile>/daemon.id — meaning the same physical machine minted a different
UUID per profile. On any host running both the CLI-spawned daemon (default
profile) and the desktop-spawned daemon (profile derived from API host),
that produced two runtime rows per provider per workspace. The server-side
`legacy_daemon_ids` merge only covers hostname variants, not UUIDs, so the
rows just piled up.
Profile boundaries are about which backend/account the daemon is talking
to, not about the physical machine. Identity should be per-machine, token
should be per-profile.
Changes:
- `EnsureDaemonID` now always reads/writes ~/.multica/daemon.id regardless
of the `profile` argument. The argument is retained for migration-only
use (see promotion below).
- Migration path: when the canonical file is missing and the requested
profile has a pre-change per-profile daemon.id, promote that UUID in
place so a user who only ever ran under a named profile keeps the same
identity instead of minting a fresh UUID and round-tripping a merge.
- New `LegacyDaemonUUIDs()` scans ~/.multica/profiles/*/daemon.id and
returns every UUID that survives parsing. `config.go` now appends those
to the daemon's `legacy_daemon_ids` payload, so any runtime rows
previously registered under a per-profile UUID (on any backend) get
merged into the canonical machine UUID at register time.
Tests replace the `ProfileIsolated` assertion with `SharedAcrossProfiles`
and add coverage for promotion, UUID scanning (including skipping corrupt
files), and the empty-profiles-dir fast path.
Adds two new toggleable card properties that surface issue context at a glance:
- Project: shows the parent project icon + title when the issue belongs to one.
- Sub-issue progress: gates the existing progress ring behind a card property
so users can hide it when not useful.
Both default to on; toggled via the existing "Display" popover.
* feat(daemon): persistent UUID identity + legacy-id merge at register-time
daemon_id is now a stable UUID persisted to `<profile-dir>/daemon.id` on
first start, replacing the hostname-derived id that drifted whenever
`.local` appeared/disappeared, a system was renamed, or a profile
switched — each of which used to mint a fresh `agent_runtime` row and
strand agents on the old one.
To migrate existing installs without operator intervention, the daemon
reports every legacy id it may have registered under previously
(`host`, `host` with `.local` stripped, and `host[-profile]` variants
for both). At register-time the server looks up each candidate row
scoped to (workspace, provider), re-points its agents and tasks onto
the new UUID-keyed row, records which legacy id was subsumed in the
new `legacy_daemon_id` column for audit, and deletes the stale row.
Result: users running `xxx.local`-keyed runtimes today transparently
land on the new UUID row on next daemon restart.
The hostname-prefix `MigrateAgentsToRuntime` / `daemon_id LIKE '...-%'`
compatibility shim is no longer needed and has been removed along with
the handler call that invoked it.
* fix(daemon): handle bidirectional .local drift and case drift in legacy merge
Review on #1220 flagged two gaps in the legacy-id migration candidate set:
1. Reverse .local: LegacyDaemonIDs only added the stripped variant when the
current hostname ended in `.local`. The opposite direction — DB has
`foo.local`, current host is `foo` — was missed, so runtimes registered
under the `.local` variant stayed orphaned after upgrade. Now both
variants (`foo` and `foo.local`) are always emitted, regardless of what
`os.Hostname()` currently returns, plus their `-<profile>` suffix forms.
2. Case drift: os.Hostname() has been observed returning different casings
on the same machine across mDNS/reboot state. A case-sensitive `=`
comparison stranded rows like `Jiayuans-MacBook-Pro.local` when the
daemon later reported `jiayuans-macbook-pro.local`. FindLegacyRuntimeByDaemonID
now uses `LOWER(daemon_id) = LOWER(@daemon_id)` on both sides, so casing
differences merge rather than orphan. The (workspace_id, provider) prefix
still bounds the scan to a tiny set of rows so the non-indexed LOWER()
comparison has negligible cost.
Tests: TestLegacyDaemonIDs gets the mixed-case + reverse-direction cases;
daemon_test.go adds TestDaemonRegister_MergesLegacyDaemonIDRuntime_ReverseDotLocal
and TestDaemonRegister_MergesLegacyDaemonIDRuntime_CaseDrift.
* fix(daemon): consolidate every case-duplicate legacy runtime, not just the first
Follow-up review on #1220: after switching to `LOWER(daemon_id) =
LOWER(@daemon_id)`, the single-row lookup still only merged one legacy
row per candidate. If a machine already had two rows in the DB that
differed only in casing (e.g. `Jiayuans-MacBook-Pro.local` AND
`jiayuans-macbook-pro.local` coexisting because earlier hostname drift
already minted a duplicate), only one of them got consolidated and the
other stayed orphaned — violating the "no duplicate runtime per machine
after backfill" acceptance.
- FindLegacyRuntimeByDaemonID → FindLegacyRuntimesByDaemonID (:many)
- mergeLegacyRuntimes iterates every returned row and dedupes across
overlapping legacy candidates so `foo` and `foo.local` both resolving
to the same stored row don't double-process
Test: TestDaemonRegister_MergesAllCaseDuplicateLegacyRuntimes seeds two
case-duplicate rows with one agent each and confirms both rows are
deleted and both agents end up on the new UUID-keyed row.
These trigger kinds exist in the DB schema but nothing on the server
fires them:
- autopilot_scheduler.ClaimDueScheduleTriggers filters kind='schedule'
(pkg/db/queries/autopilot.sql:150)
- DispatchAutopilot is reached only from the scheduler (source:schedule)
or POST /api/autopilots/{id}/trigger (source:manual); no inbound
webhook or api endpoint exists
- The UI only surfaces schedule creation
Exposing them in the CLI lets users create triggers that sit in the DB
doing nothing. Drop --kind from trigger-add, require --cron, always
send kind=schedule. Re-add the flag when the server grows a dispatch
path for the other kinds.
Follow-up to #1249. Two small follow-ups requested in review:
1. `resolveTaskWorkspaceID` was duplicated between `handler/daemon.go` and
`service/task.go`. #1249 fixed the handler copy but left both in place,
meaning any future branch (e.g. a fourth task link type) still needs
to be added in two files. Promote the service method to the exported
`TaskService.ResolveTaskWorkspaceID` and delete the handler copy.
Handler's `requireDaemonTaskAccess` and `ListTaskMessagesByUser` now
call through `h.TaskService`.
2. Add a regression test `TestStartTask_AutopilotRunOnlyTask_ResolvesWorkspace`
covering the exact scenario from #1224: a task linked only via
`AutopilotRunID` must resolve to the autopilot's workspace. The test
asserts 404 for a cross-workspace daemon token and 200 (with status
transitioning to `running`) for the correct-workspace token.
Follow-up to #1192. Document the v2 protocol contract that the
dispatch-level threadId guard relies on, and lock down the two leakage
paths the guard closes:
- turn/completed from a subagent thread must not call onTurnDone
- item/completed (agentMessage, final_answer) from a subagent thread
must neither leak text into the output builder nor terminate the turn
Without these tests a future refactor that drops or relocates the guard
would not be caught by CI, since existing notification tests omit the
top-level threadId field and pass through unfiltered.
* feat(cli): add autopilot commands
Expose the existing autopilot REST API through the multica CLI so
users and agents can list, get, create, update, delete, trigger, and
inspect autopilots, plus manage their triggers (schedule/webhook/api).
Also surface the read + core write commands in the agent meta skill
prompt so agents discover them without needing --help.
- new cmd_autopilot.go (+ test) wiring /api/autopilots endpoints
- add APIClient.PatchJSON (autopilot update uses PATCH)
- expose autopilot in CORE COMMANDS group
- extend runtime_config.go meta skill with autopilot entries
- document autopilot command group in CLI_AND_DAEMON.md
* fix(autopilot): address code review — restrict run_only, validate workspace on update
Code review caught two issues with the initial CLI PR:
1. run_only mode is broken end-to-end. The daemon-side
resolveTaskWorkspaceID() in internal/handler/daemon.go only resolves
workspace from issue/chat, so run_only tasks (which have neither)
return 404 from /start. BuildPrompt() would also emit an empty issue
ID. The service-level resolver in internal/service/task.go already
handles AutopilotRunID, but the daemon endpoint uses the handler
copy. Fixing that path is out of scope for the CLI PR; drop
run_only from the CLI and docs so we don't recommend a mode that
cannot complete. Server continues to accept it for the existing UI.
2. UpdateAutopilot did not verify that a new assignee_id belongs to
the workspace, unlike CreateAutopilot. This let a PATCH swap in an
agent from a different workspace. Mirror the same
GetAgentInWorkspace check.
The route-level loading.tsx creates a Suspense boundary that shows a
generic skeleton on every page navigation within the dashboard. Since
every page already handles its own data-loading skeleton via TanStack
Query isLoading, this causes two sequential skeleton flashes:
loading.tsx skeleton → page skeleton → content.
Removing it makes the old page stay visible during route transitions
(typically <100ms), then the new page renders directly with its own
skeleton — a single, smooth transition.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(daemon): filter thread/status/changed by threadId to prevent subagent interference
When Codex CLI has memories enabled, the app-server spawns a memory
consolidation subagent as a separate thread within the same stdio
connection. When that subagent thread finishes and transitions to idle,
the daemon's codex backend mistakenly interprets the idle signal as the
main turn completing, causing it to close stdin and cancel the context
before the real turn produces any output.
Add a threadId check to the thread/status/changed handler so only
status changes from the tracked thread trigger turn completion. Signals
from subagent threads (threadId != c.threadID) are now ignored.
Fixes#1181
* fix(codex): dispatch-level threadId filter for subagent notifications
Codex multiplexes subagent threads (e.g. memory consolidation) on
the same stdio pipe. Previously only thread/status/changed had a
threadId guard, but item/completed (agentMessage + final_answer),
turn/completed, and turn/started from subagent threads could still
trigger onTurnDone or contaminate output.
Move the threadId check to the top of handleRawNotification so all
notification handlers are protected. Remove the now-redundant
per-handler check on thread/status/changed.
Fixesmultica-ai/multica#1181
---------
Co-authored-by: fuxiao <fuxiao@zyql.com>
resolveTaskWorkspaceID only handled tasks linked via IssueID or
ChatSessionID. Tasks created by run_only autopilots (introduced in
#1028) have only AutopilotRunID set, so the resolver returned an empty
workspace ID, causing requireDaemonTaskAccess to respond with 404.
Add an AutopilotRunID branch that looks up the autopilot run, then
its parent autopilot, to obtain the workspace ID.
The project CRUD commands (list, get, create, update, delete, status)
and the `--project` flag on issue commands have been implemented in
the CLI but were not yet documented. Add them to both the docs site
reference and the repo-level CLI_AND_DAEMON.md so the feature is
discoverable.
Closes MUL-867
Co-authored-by: Eve <eve@multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Make the http/https scheme allowlist structurally enforced instead of a
convention. Move the allowlist check + shell.openExternal call into a
single openExternalSafely wrapper in external-url.ts, have both main-process
call sites (the IPC handler and setWindowOpenHandler) go through it, and
add an ESLint no-restricted-syntax rule that bans direct shell.openExternal
usage anywhere under apps/desktop/src/main/ except external-url.ts itself.
This is the follow-up to #1124: same safety guarantee, but a reviewer can
no longer accidentally reintroduce a bare shell.openExternal somewhere that
bypasses the check — the lint rule catches it at CI time. Also restores
the scheme info in the warn log (lost when the helper was extracted).
Test coverage extended to the cases the original PR review flagged but
didn't ship: casing (FILE:// / HTTPS://), javascript: / data:, ftp / smb,
vscode:// / ms-msdt:, mailto / tel, credentials-in-URL, empty / malformed.
Added two openExternalSafely tests (electron mocked) confirming allowed
URLs forward and rejected URLs do not.
Closes a follow-up bullet from the internal #1115 / #1124 review.
* fix(desktop): restrict shell.openExternal to http/https schemes
The Electron main-process IPC handler for shell:openExternal called
shell.openExternal with whatever string the renderer passed, with no
scheme validation. Under this app's intentional webSecurity: false and
sandbox: false configuration (#648), any unsafe content path in the
renderer reaching this IPC becomes a way to dispatch arbitrary OS
protocol handlers — file://, smb://, vscode://, Windows ms-msdt:,
and so on.
Parse the URL and reject anything outside http/https (the only schemes
any legitimate call site uses today). Matches the Electron security
checklist guidance for openExternal on non-isolated renderers.
Closes#1115
* Close the desktop external-open gap on target=_blank links
The original fix validated only the IPC path, but the renderer could still trigger shell.openExternal through setWindowOpenHandler for target="_blank" links and window.open(). This change reuses one allowlist helper for both sinks and adds a focused unit test for the helper contract.
Constraint: Desktop shell.openExternal must stay limited to http/https despite webSecurity=false and sandbox=false
Rejected: Duplicate URL validation logic in each sink | easy to drift and harder to test
Confidence: high
Scope-risk: narrow
Reversibility: clean
Directive: Keep all desktop external-open paths on the same validator so new sinks do not bypass the allowlist
Tested: pnpm --dir /Users/jh0927/Workspace/multica-pr1124-followup --filter @multica/desktop test -- src/main/external-url.test.ts
Tested: pnpm --dir /Users/jh0927/Workspace/multica-pr1124-followup --filter @multica/desktop typecheck
Not-tested: Full desktop app manual smoke run
Related: #1115
---------
Co-authored-by: shaun0927 <shaun0927@users.noreply.github.com>
* fix(daemon): platform-aware Codex sandbox config to unbreak macOS network
On macOS, Codex's Seatbelt sandbox in workspace-write mode silently
ignores '[sandbox_workspace_write] network_access = true' (see
openai/codex#10390). That blocks DNS inside the sandbox, so 'multica
issue get' and other CLI calls fail with 'dial tcp: lookup ...: no such
host' — this is what caused MUL-963.
Changes:
- New server/internal/daemon/execenv/codex_sandbox.go: picks a sandbox
policy based on runtime.GOOS and the detected Codex CLI version.
Non-darwin or darwin with a known-fixed version keeps workspace-write
+ network_access=true; older darwin falls back to danger-full-access
and logs a warn with upgrade hint. The fix-version threshold is a
single constant (CodexDarwinNetworkAccessFixedVersion) so it's easy
to bump once upstream ships.
- Per-task config.toml now gets a 'multica-managed' marker block
(BEGIN/END comments) rewritten idempotently; user-owned keys outside
the markers are preserved. Legacy inline sandbox directives from
earlier daemon versions are stripped on migration.
- execenv.PrepareParams gains CodexVersion; execenv.Reuse takes a
codexVersion arg; daemon.go caches detected versions at registration
and threads them through to Prepare/Reuse.
- Replaces the old ensureCodexNetworkAccess tests with
platform-parameterised coverage (linux vs darwin, idempotency,
legacy-migration, policy matrix).
- docs/codex-sandbox-troubleshooting.md: symptom fingerprint table,
decision matrix, self-check commands, trade-offs.
Refs: MUL-963
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(daemon): hoist managed sandbox block above user tables (MUL-963)
Review on #1246 flagged that upsertMulticaManagedBlock appended the
managed block to EOF. If the user's config.toml ends inside a TOML table
(e.g. [permissions.multica] or [profiles.foo]), a trailing bare
sandbox_mode = "..." is parsed as a key of that preceding table, so
Codex silently ignores the policy the daemon meant to apply.
Two changes make the block position-independent:
- renderMulticaManagedBlock now emits only top-level key=value lines and
uses TOML dotted-key form (sandbox_workspace_write.network_access =
true) instead of opening a [sandbox_workspace_write] header. The block
therefore neither inherits from nor leaks into any surrounding table.
- upsertMulticaManagedBlock always hoists the block to the top of the
file (stripping any previously written managed block first), so the
sandbox_mode line is always at the TOML root regardless of what the
user put below it. This also migrates configs written by the original
PR #1246 logic where the block was trapped behind a user table.
Added tests for the regression scenario (pre-existing [permissions.*]
table) and the legacy-trailing-block migration; updated the existing
Linux default test and the troubleshooting runbook to reflect the
dotted-key form.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: CC-Girl <cc-girl@multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Autopilot was formatting the triggered-at timestamp with time.RFC3339
(e.g. "2026-04-16T14:54:32Z"), which is hard to read and confusing for
users in non-UTC timezones because the "Z" suffix looks like an error
instead of a timezone indicator.
Switch to a human-readable format ("2026-04-16 14:54 UTC") so only the
hour differs from local time; minutes match across timezones, making
the value easy to reconcile at a glance.
Fixesmultica-ai/multica#1197.
Shared inbox links (?issue=<id>) pointed to notifications that may no
longer exist in the current user's inbox (archived, or received by
someone else). The detail pane would fall back to an empty state and
leave the user stuck.
After inbox loads, if the selected key has no matching item, replace
the URL with /issues/<id> so the link still resolves to something
meaningful.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
crypto.randomUUID() is only defined in secure contexts, so self-hosted
HTTP deployments were throwing TypeError on mount and when clicking Add.
Route the id generation through the existing createSafeId() helper so
the tab works in non-secure contexts too.
Fixes#1214
- Migration 046 adds UNIQUE(workspace_id, name) with dedup (keep most recently updated)
- CreateAgent handler returns 409 Conflict scoped to constraint name agent_workspace_name_unique
- Dedup verified as (0 rows) against worktree DB; rerun against staging/production before applying
- Down migration drops the constraint only; deleted rows and cascaded data are not restored
Co-authored-by: Anup Joy <joyanup@gmail.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Desktop-launched daemons have their CLI binary overwritten by the Desktop
app on every launch, so any in-app update is reset. The detail panel already
renders 'Managed by Desktop' and hides the Update button when
metadata.launched_by === 'desktop', but the sidebar red dot
(useMyRuntimesNeedUpdate) and the list arrow (useUpdatableRuntimeIds) still
flagged them because runtimeNeedsUpdate() only considered mode/owner/version.
Short-circuit runtimeNeedsUpdate() on launched_by === 'desktop' so all three
surfaces (sidebar dot, list arrow, detail panel) agree and defer CLI
upgrades to the Desktop auto-updater.
Co-authored-by: CC-Girl <cc-girl@multica.ai>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Our CLI release flow pre-creates a *published* GitHub Release via
`gh release create`. electron-builder's default `publishingType: draft`
conflicts with `existingType=release` and causes the DMG/ZIP/blockmaps/
latest-mac.yml uploads to be silently skipped, which breaks
electron-updater auto-update on installed clients (observed on v0.2.4,
had to fall back to `gh release upload` manually).
Explicitly setting `publishingType: release` aligns electron-builder
with our release flow so desktop artifacts are uploaded to the existing
published release automatically.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* refactor(desktop): tabs are per-workspace, not cross-workspace
Tabs are now grouped by workspace in the store; the TabBar shows only the
active workspace's tabs, and switching workspace swaps the visible group.
Before this change tabs were a flat list that spanned workspaces, which
produced a confusing experience: working in acme with three tabs, then
switching to butter and back, still showed whatever tabs you happened to
open while you were in butter alongside your acme work.
The bug had the same shape as the pre-workspace-overlay bug we fixed in
#1237 — a concept ("workspace") was encoded in data (tab paths) but
ignored by the UI that displayed it (TabBar). The fix is structural:
make the data model match the concept.
Key changes:
- **Schema**: `{ activeWorkspaceSlug, byWorkspace: {slug: {tabs, activeTabId}} }`.
The invariant "every tab belongs to a workspace group" is enforced at
sanitize time and at migration time; there is no longer a root `/`
sentinel.
- **NavigationAdapter** detects cross-workspace pushes and delegates to
`switchWorkspace(slug, path)` instead of navigating the active tab's
router. All existing call sites in shared code (sidebar dropdown,
settings post-delete redirect, invite-accept, cmd+k) keep calling
`push(paths.workspace(x).issues())` unchanged.
- **TabContent** renders only the active workspace's tabs under Activity.
Cross-workspace state preservation is an explicit non-goal — switching
workspaces should feel like switching.
- **WorkspaceRouteLayout** auto-heal no longer navigates the tab router
to `/`. Stale-slug cleanup is a store-level op (`validateWorkspaceSlugs`)
that drops the whole stale group in one go.
- **App.tsx** bootstrap seeds `activeWorkspaceSlug` when null and the
user has workspaces; the new-workspace overlay opens/closes based on
workspace count independently of any route.
- **Persistence migration** (v1 → v2) groups old flat tabs by extracted
slug, drops root / transition / reserved-slug tabs, and picks an
active workspace from the old active tab's owning group. No data
loss for existing users with workspace-scoped tabs.
Web is unchanged — tabs are a desktop-only concept. `packages/views`,
`packages/core`, `apps/web` are all untouched. `setCurrentWorkspace`
in core remains the single source of truth for the API client's
workspace header, driven by `WorkspaceRouteLayout` as before.
Tests: 19 tab-store tests (sanitize, migration, switchWorkspace,
validate, close-last-reseeds, reset). 38 desktop tests total pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* review: stable selectors + defensive guards on tab-store
Addresses self-review findings on #1239.
**C1 — perf cliff from unstable selector returns.** The previous
`useActiveTab()` selector used `.find()` inside, so every router tick
on the active tab (which replaces the Tab object via immutable spread
in updateTab / updateTabHistory) forced every subscriber to re-render.
Replaced with finer-grained selectors:
- `useActiveTabIdentity()` — { slug, tabId } primitives (stable across
unrelated updates).
- `useActiveTabRouter()` — stable object reference for a tab's lifetime.
- `useActiveTabHistory()` — { historyIndex, historyLength } numbers.
`useTabHistory` and `DesktopNavigationProvider` now consume the
primitive selectors, so back/forward buttons don't churn on every
path change. A non-hook `getActiveTab(state)` helper covers the
event-handler case.
**I1 — `switchWorkspace` no-ops on empty slug.** Defensive guard in
case a malformed path ever reaches the adapter's detector.
**I2 — merge warns on path/slug mismatch.** Previously silent drop;
now `console.warn` makes the condition visible during debugging.
**Misc — TabRouterInner takes `tab` prop directly.** Passing the Tab
object eliminates a redundant store read per rendered tab.
Known follow-up (not this PR): `packages/core/realtime/use-realtime-sync.ts`
still uses `window.location.assign` for workspace-deleted eviction —
that's a full renderer reload on desktop, which post-refactor wastes
the careful in-memory tab state we just set up. Fixing cleanly requires
a navigation-callback injection pattern through CoreProvider, which is
cross-cutting and deserves its own PR.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(workspace): navigate away BEFORE leave/delete mutation to avoid CancelledError
Symptom: deleting the current workspace logged "current workspace
deleted, switching" from the realtime handler and surfaced an
"Uncaught (in promise) CancelledError" from TanStack Query's
refetchQueries batch.
Root cause: a three-way race between the mutation's own
invalidateQueries(workspaceKeys.list()), the settings page's
navigateAwayFromCurrentWorkspace() fetchQuery, and the realtime
workspace:deleted handler's relocateAfterWorkspaceLoss fetchQuery.
All three refetched the same query concurrently; TanStack Query
cancelled the in-flight loser(s), and the rejection bubbled out of
invalidateQueries as an unhandled promise rejection.
Fix: invert the order. Compute the destination from the current
cached workspace list, navigate immediately, *then* fire the
mutation. By the time the backend fires workspace:deleted, the
active workspace is already something else — the realtime handler's
"current === deleted" check fails and its relocate branch no-ops.
Only one refetch happens (the mutation's onSettled), no race, no
cancellation.
navigateAwayFromCurrentWorkspace no longer needs async/fetchQuery
since it reads from cache and returns before the mutation fires.
Applies to both Leave and Delete flows. Both web and desktop benefit
since the code is in packages/views.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(desktop): clear workspace singleton + flex drag strip + defer seeding
Three issues that the last round of delete-workspace fixes missed.
**1. `setCurrentWorkspace` singleton leaks after delete.** Navigating
before the mutation (prior fix) changed the URL but nothing cleared
the core platform's currentSlug/currentWsId singleton. Three
downstream consumers still believed the deleted workspace was active:
- `useRealtimeSync`'s `workspace:deleted` handler: its
`getCurrentWsId() === deleted` check fired, triggering a parallel
relocate that raced the mutation's invalidate and the settings
page's navigate — CancelledError + `window.location.assign`
(white screen reload).
- Chrome gating: `{slug && <AppSidebar />}` stayed truthy, the
sidebar mounted, and `useWorkspaceId` inside it threw because the
workspace was gone from the list cache.
- API client's `X-Workspace-Slug` header: stale on the next call.
Fix: `navigateAwayFromCurrentWorkspace` now calls
`setCurrentWorkspace(null, null)` before pushing. The next
workspace's `WorkspaceRouteLayout` re-sets the singleton when it
mounts; for the last-workspace case, null is the correct state
(overlay has no workspace context).
Same family as the previous logout bug: persist only writes to
storage, reset on logout must also wipe in-memory state. Here the
singleton is another in-memory bit that survives a URL change if
we don't explicitly clear it.
**2. "Cannot update a component while rendering" warning.** The
per-workspace-tabs refactor kept the validate+seed call in render
phase (matching the pre-refactor pattern). It worked before because
`validateWorkspaceSlugs` is idempotent; the new `switchWorkspace`
seed is not, and triggers a TabBar re-render during AppContent's
render. Moved to `useLayoutEffect` — synchronously after render,
before paint, no flicker.
**3. Welcome-screen drag region didn't work on desktop.** The
absolute-positioned `h-10 z-10` drag strip relied on z-index stacking
to beat the content wrapper's no-drag for hit-testing, which wasn't
reliable for `-webkit-app-region` on the overlay. Replaced with a
flex child (`h-12 shrink-0` at top of the overlay's flex-col), so
the drag region owns its own layout space — any pixel in the top 48
is unambiguously drag.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* docs(CLAUDE): desktop-specific rules — routing, singleton, drag, UX split
Codifies the lessons from the recent desktop refactor series
(#1237, #1238, #1239) so future work doesn't re-derive them from
bugs. Covers:
- **Route categories** (session / transition / error) — explains why
`/workspaces/new` and `/invite/:id` are overlay state, not routes,
on desktop; stale slugs auto-heal instead of rendering error pages.
- **`setCurrentWorkspace` singleton hygiene** — unmount doesn't
clear it; any code leaving workspace context must call
`setCurrentWorkspace(null, null)` explicitly.
- **Workspace destructive operations ordering** — navigate first,
mutate after, to avoid the three-way refetch race that surfaces
as CancelledError + full-page reload.
- **Tab isolation** — tabs are grouped per workspace; cross-workspace
push is intercepted by the navigation adapter and translated into
switchWorkspace.
- **Drag region pattern** — flex child at top, not absolute overlay;
`-webkit-app-region` hit-testing is unreliable with z-index stacking.
- **UX vs platform chrome split** — UX affordances (Back, Log out,
welcome copy) in packages/views/; platform chrome (drag, immersive
mode, tab system) in desktop-only code.
Also patches the Cross-Platform Development Rules' rule #2 which
previously said "add a route in both apps" unconditionally — added
the exception for pre-workspace transition flows pointing at the
new Desktop-specific Rules section.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(editor): prevent duplicate image attachments showing as file cards
Images pasted into comments could produce duplicate attachment records
(macOS/Chrome clipboard provides duplicate File entries), causing
AttachmentList to show a spurious file card below the inline image.
Three-layer fix:
- Dedup clipboard files by name+size+type in paste/drop handlers
- Track upload URL→ID mapping instead of accumulating IDs blindly;
only send IDs for uploads still present in content on submit
- AttachmentList filters duplicate attachments (same file identity)
where a sibling is already referenced inline — handles old data
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(editor): rewrite bubble menu with @floating-ui/dom for reliable scroll hiding
Replace Tiptap's native <BubbleMenu> with custom @floating-ui/dom positioning.
The native plugin's virtual element lacked contextElement, so the hide middleware
could only detect viewport clipping — not nested scroll container clipping
(e.g. comment/reply inputs inside a scrollable page).
Key changes:
- contextElement on virtual reference enables hide middleware to detect ALL
overflow ancestor clipping, not just viewport bounds
- visibility:hidden (not display:none) keeps element measurable for
computePosition, fixing the comma-selection positioning bug
- autoUpdate monitors all scroll ancestors automatically via contextElement
- Remove: getScrollParent, scrollHiddenRef, manual scroll listener, scrollTarget
- Remove @tiptap/extension-bubble-menu dependency (no longer used)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(workspace): typed delete confirm + sole-owner leave preflight
Harden the Danger Zone on the Workspace settings tab.
**Delete workspace** now requires typing the workspace name exactly
(case-sensitive, no trimming) before the destructive button enables —
GitHub's repo-delete pattern. Deleting cascades into every issue,
agent, skill, and run under the workspace with no soft-delete, so the
friction is deliberate. Enter submits only when matched; the input
clears on close so reopening for a different workspace doesn't leak
the prior attempt.
**Leave workspace** now preflights the sole-owner case the backend
already blocks (server/internal/handler/workspace.go:569 — "workspace
must have at least one owner"). Previously the user clicked Confirm
and got an opaque 400 toast; now the Leave button is disabled upfront
with inline guidance that distinguishes:
- sole member: "Delete the workspace to leave."
- sole owner with other members: "Promote another member to owner
first, or delete the workspace."
Both changes live in packages/views/, so web and desktop get the same
Danger Zone treatment automatically.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* review: gate Danger Zone on members fetched + reset typed input on rename
Addresses self-review findings on #1238:
- Previously the Danger Zone rendered immediately with `members = []`, so
the Delete workspace block (gated on `isOwner`, which is derived from
an empty members list) would flash in once the query settled. Gate the
whole section on `membersFetched` so it appears once with correct
controls.
- Reset `typed` on `workspaceName` change too — if another owner renames
the workspace while the dialog is open, the already-typed string stops
matching silently; resetting surfaces the mismatch.
- Added two tests: unicode/special-char names match literally; rename
mid-dialog clears the input.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Previously /workspaces/new and /invite/:id were tab routes on desktop.
That meant the TabBar rendered on top of flows that conceptually aren't
"places" the user sits at — creating a workspace or accepting an invite
is a one-shot transition, not a session. The mismatch also produced
several downstream bugs: tab state persisted these paths, the invite
deep link had no clean dispatch target, and NoAccessPage leaked TabBar
chrome when a workspace slug went stale.
Fix by recognising the underlying category mistake: on desktop, these
flows are application state, not routes. Move them to a window-level
overlay driven by a small Zustand store; the navigation adapter
intercepts pushes to the corresponding paths and routes them to the
overlay instead. Web keeps the routes (users need shareable URLs and
back-button semantics), so shared view components are reused as-is.
UX affordances (Back button when dismissable, Log out escape) live in
the shared NewWorkspacePage/InvitePage so both platforms render
identical content; the desktop overlay is now a thin platform shell
(drag strip + useImmersiveMode) that wraps the shared UX. Web wires
onBack based on whether the user has any workspaces.
Also addresses several related issues uncovered along the way:
- Logout now resets the in-memory tab + overlay stores (previously only
localStorage was cleared, so the next login inherited the prior
user's tabs).
- WorkspaceRouteLayout auto-heals a stale workspace slug by navigating
to "/" instead of rendering NoAccessPage — on desktop without a URL
bar, "no access" is always stale state, not a legitimate destination.
- IndexRedirect overlay lifecycle is bidirectional: opens when wsList
is empty, closes when it becomes non-empty (realtime workspace:added
would otherwise leave the overlay stuck open).
- tryRouteToOverlay resets the current tab to "/" when opening the
new-workspace overlay; otherwise workspace-scoped components under
the overlay continue to render and throw when the workspace they
reference disappears from the cache (reproducible by deleting the
last workspace from Settings).
- handleDeepLink now accepts multica://invite/<id>, IPC'd through to
the renderer and opened as an invite overlay. Email template still
links to https:// (unchanged), but the desktop dispatch path is now
wired for a future "open in desktop app" bridge.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Images pasted into comments could produce duplicate attachment records
(macOS/Chrome clipboard provides duplicate File entries), causing
AttachmentList to show a spurious file card below the inline image.
Three-layer fix:
- Dedup clipboard files by name+size+type in paste/drop handlers
- Track upload URL→ID mapping instead of accumulating IDs blindly;
only send IDs for uploads still present in content on submit
- AttachmentList filters duplicate attachments (same file identity)
where a sibling is already referenced inline — handles old data
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The issue:created subscriber listener type-asserted payload["issue"] to
handler.IssueResponse, but autopilot publishes the issue as
map[string]any (via service.issueToMap). The assertion failed silently,
so no subscribers (including the creator) were ever added to autopilot
issues — meaning creators received no notifications when their
autopilot run produced comments or status changes.
Add an extractIssueFields helper that accepts either format and use it
in both the issue:created and issue:updated listeners. Mirrors the
dual-format pattern already used by the comment:created listener.
Replace the Pause/Activate button on the detail page with a Switch next
to the title, showing a colored status label. Flipping it toggles
between active and paused via the existing updateAutopilot mutation.
* refactor(landing): tighten hero — CTAs, install copy, works-with wrap, LCP priority
- Drop GitHub button from hero CTAs (already in header) so the primary
Start / Download Desktop pair is the clear path.
- Split InstallCommand: outer is no longer a <button>, so text selection
no longer fights with copy. Mobile gets full-width with break-all;
desktop keeps the compact pill. Copy button has aria-label.
- Fix invalid `hover:bg-white/8` opacity to `hover:bg-white/[0.08]` so
the install pill's hover background actually renders.
- Add `flex-wrap` and gap-y to the "Works with" row so the label + 5
logos can stack on small screens instead of overflowing horizontally.
- Move `priority` from the decorative backdrop image onto the product
hero image (the actual LCP candidate) to stop background bytes from
starving the foreground.
* refactor(landing): remove install command from hero
Per design feedback, the install command pill is removed from the hero.
The download path now flows through the Download Desktop CTA only;
install instructions remain available in the docs and README.
Previously every registered Command (New Issue, New Project, three theme
switches, plus contextual Copy actions on issue pages) surfaced on empty
query, leaving only 3–5 rows for Recent in a 400px panel. Low-frequency
commands (theme, copy, New Project) are now revealed by typing, matching
the progressive-disclosure pattern already used for Pages and Switch
Workspace. Refs MUL-991.
* feat(search): add light/dark/system theme toggle actions to cmd+k
The command palette now surfaces an "Actions" section with theme toggle
items (Light / Dark / System), searchable via keywords like "theme",
"light", "dark", "appearance", or "mode". The active theme is marked
with a check icon.
* feat(search): add quick-win commands to cmd+k palette
Extends the command palette with a "Commands" group that consolidates
theme toggles plus four new actions:
- New Issue / New Project — trigger the global create modals
- Copy Issue Link / Copy Identifier (MUL-xxx) — only when the current
route is an issue detail page; mirrors the copy-link dropdown logic
from issue-detail
Adds a "Switch Workspace" group that lists the user's other workspaces
(filtered by name/slug, or by typing "workspace"/"switch") and
navigates to the selected workspace's issues page.
To make "New Project" work from anywhere, the inline CreateProjectDialog
on ProjectsPage is extracted into a global CreateProjectModal mounted
via the existing ModalRegistry + modal store (same pattern as
create-issue / create-workspace). The modal store type gains a
"create-project" variant.
* feat(search): show Commands by default so they're discoverable
Before, cmd+k actions (New Issue / New Project / Copy link / Copy ID /
theme toggles) only appeared when the user typed a matching keyword,
leaving them invisible unless the user already knew they existed.
Now the Commands group renders as soon as the palette opens (no query),
with the whole command list shown; typing narrows it down as before.
Also trims the redundant "⌘K to open this anytime" hint from the empty
state — the palette is already open.
Dev Electron uses a single userData path ("Multica Canary") derived from
the app name, which also locates the single-instance lock. Two worktrees
running dev simultaneously fight for that lock — the second `app.quit()`s
silently before opening a window.
DESKTOP_APP_SUFFIX appends to the app name + userData path so each
worktree can claim its own lock:
DESKTOP_APP_SUFFIX=foo → "Multica Canary foo"
Default (no env var) keeps behavior unchanged.
Complements the existing DESKTOP_RENDERER_PORT env from #1210 so a full
"run a second dev Electron" setup looks like:
DESKTOP_RENDERER_PORT=15173 DESKTOP_APP_SUFFIX=foo pnpm dev:desktop
Hooks recordVisit into useCreateIssue onSuccess so issues the user just
created appear in cmd+k's Recent section without requiring them to open
the issue first.
* feat(desktop): brand dev build as Multica Canary with bundled icon
pnpm dev:desktop ran under the stock Electron name and default icon,
making it indistinguishable from any other Electron dev app in the dock.
Set a Canary app name + userData path and point the macOS dock icon and
BrowserWindow icon at the bundled resources/icon.png so the dev build is
visually branded.
* feat(desktop): allow overriding renderer port via DESKTOP_RENDERER_PORT
Lets a second worktree run `pnpm dev:desktop` while a primary checkout
already holds the default Vite dev port 5173 — required to actually
exercise the "Multica Canary" branding in isolation.
* feat(desktop): rebrand Electron.app Info.plist so dev shows Multica Canary
app.setName() can't override the macOS menu bar title or Cmd+Tab label
— those come from CFBundleName baked into the running bundle's
Info.plist. Patch the bundled Electron.app's plist during `pnpm
dev:desktop` so dev launches read "Multica Canary" everywhere, not
"Electron". Idempotent; unlinks before rewriting so we don't mutate a
pnpm-store inode shared with other projects.
Previously shouldEnqueueOnComment suppressed agent triggers on done/
cancelled issues, requiring an explicit @mention to resume the
conversation. The gate was non-obvious and confused users who expected
a regular reply to wake the agent up.
Drop the status check — comments are conversational and should wake
the agent up at any status. @mention already bypasses all gates, so
behavior for mentions is unchanged.
Refs multica-ai/multica#1205
* feat(issues): persist comment collapse state across page reloads
Store collapsed comment IDs in a workspace-scoped Zustand store backed
by localStorage, replacing the transient useState(true) default.
Comments now remember their collapsed/expanded state per issue.
* test(issues): add useCommentCollapseStore mock to issue-detail tests
The existing vi.mock for @multica/core/issues/stores didn't include the
newly exported useCommentCollapseStore, causing CommentCard to throw at
render time.
* fix(daemon): normalize hostname by stripping .local mDNS suffix
Daemons started via different methods (standalone CLI vs desktop app
bundled binary) resolve the hostname differently on macOS — one gets
'computer' and the other 'computer.local'. This caused duplicate runtime
registrations for the same machine.
Stripping the .local suffix at the point of hostname resolution ensures
both always register under the same identifier.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* fix(daemon): move empty-host fallback to after .local trim; fix Makefile @ prefix
- Reorder: TrimSuffix runs first, then empty-check, so a hostname of
just ".local" doesn't propagate as an empty daemon_id/device_name
- Add missing @ prefix on migrate command in Makefile so it isn't
echoed twice at startup
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
When invoked as `pnpm package -- --mac --arm64 --publish always`,
the bare `--` separator that pnpm inserts was forwarded into
electron-builder's argv. This terminated option parsing, causing
`--publish always` to be treated as positional arguments instead of
a named flag. As a result electron-builder built locally but never
uploaded artifacts to the GitHub Release (isPublish: false).
Add `stripLeadingSeparator()` to remove the leading `--` before
passing args through. Includes unit tests.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(desktop): new tab inherits current workspace + guard against malformed tab paths
Three layered fixes for the same root cause: tab URLs were being
constructed without a workspace slug in some code paths, triggering
NoAccessPage whenever the router interpreted the first segment as a
(non-existent) workspace slug.
## Layer 1 — tab-bar "+" button now inherits current workspace
The handler had a hardcoded `path = "/issues"` left over from before
the slug URL refactor. Without a workspace prefix, the router saw
`workspaceSlug = "issues"` and rendered NoAccessPage. Read
`getCurrentSlug()` and build `/{slug}/issues` instead. Falls back to
"/" (→ IndexRedirect) when there is no current workspace.
This matches terminal/IDE new-tab semantics: new tab opens in the
same workspace as the active tab, not in `wsList[0]`.
## Layer 2 — validateWorkspaceSlugs runs synchronously
PR #1178 added startup validation of persisted tab slugs against the
current workspace list, but ran it in a useEffect. useEffect fires
AFTER commit, so the initial render would briefly show NoAccessPage
on a stale slug before the effect reset the tab path. Moving the call
into render phase eliminates that flash; zustand supports setState in
render, and the validator is idempotent (early-returns if nothing
changed) so this doesn't loop.
## Layer 3 — tab store rejects malformed paths at construction
Any path whose first segment is a reserved slug (e.g. "/issues",
"/login") clearly lacks a workspace prefix and is a caller bug.
sanitizeTabPath catches these at makeTab time, rewrites to "/", and
logs a console.warn naming the offending path so the bug can be fixed
at source. Any future new-tab entry point that forgets the slug will
not reach NoAccessPage.
Net effect: NoAccessPage is reserved for its legitimate purpose —
users navigating to URLs they genuinely don't have access to — and
can no longer be triggered by system bugs.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* review: read new-tab workspace from active tab + unify sanitize + add tests
Three follow-ups from self-review of PR #1198:
1. Resolve the current workspace from the active tab's path instead of
from getCurrentSlug(). With N tabs mounted under <Activity>, every
WorkspaceRouteLayout calls setCurrentWorkspace() in render — the
singleton ends up holding "whichever tab rendered last", which is
non-deterministic. activeTabId is the unambiguous source of truth
for "which workspace is the user actually looking at right now".
2. Unify the persist merge's stale-path detection with sanitizeTabPath.
The merge previously checked ROUTE_ICONS (dashboard segments only);
sanitizeTabPath uses isReservedSlug (dashboard + auth + platform +
RFC 2142 + hostname confusables). Same code path now, wider
coverage, and one source of truth.
3. Add unit tests for sanitizeTabPath: root pass-through, global paths,
valid workspace-scoped paths, malformed paths (reserved first
segment) rejected with console.warn, and user slugs that happen to
look path-like but aren't reserved.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(routing): rename /new-workspace to /workspaces/new + extend reserved slug list
Two related changes:
1. Rename the global workspace-creation route from /new-workspace to
/workspaces/new. The hyphenated word-group `new-workspace` is a
common user workspace name (last deploy was blocked by a real user
with exactly this slug). Industry consensus from auditing Linear,
Vercel, Notion, Slack, GitHub: zero major SaaS uses hyphenated
word-group root routes — they all use single words or `/{noun}/{verb}`
pairs. Reserving the noun `workspaces` automatically protects the
entire `/workspaces/*` subtree, so future workspace-related routes
(`/workspaces/{id}/edit`, `/workspaces/{id}/billing`, etc.) need no
additional reserved slugs or audit migrations.
2. Extend the reserved slug list to cover the minimal set recommended by
the URL-design audit: full auth flow vocab, RFC 2142 mailbox names
(postmaster, abuse, noreply...), hostname confusables (mail, ftp,
static, cdn...), and likely-future platform routes (docs, support,
status, legal, privacy, terms, security, etc.). Production data
audit confirmed zero conflicts for every newly added slug, so
migration 047 (the safety net) passes cleanly.
Slugs intentionally NOT added despite being in scope of the audit:
admin, multica, new, setup, www. Each has one production workspace
already using it; adding them now would block deploy. They will be
handled in a follow-up PR via owner outreach + targeted rename.
Also adds a CLAUDE.md convention rule: new global routes MUST use a
single word or `/{noun}/{verb}` pair, never hyphenated word groups.
This prevents the pattern from regenerating itself.
This PR does NOT resolve the currently-blocked prd deploy — that requires
the existing `slug='new-workspace'` workspace (owner: Dhruv Raina) to be
renamed by ops. After that workspace is renamed and migration 046 passes,
this PR's migration 047 will also pass on its first run.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* review: drop migration 046, sweep stale comments, drive reserved test from map
Address code review on PR #1188:
1. Delete migration 046 (audit_new_workspace_slug). It audits "new-workspace"
which is no longer a reserved slug after this PR's rename. Removing 046
has an unexpected upside: it directly unblocks the currently-stuck prd
deploy. Migration 046 had never successfully applied (it was the source
of the deploy block); the audit-only nature means down-rollback is a
no-op. The user workspace previously caught by 046 (slug='new-workspace',
owner: Dhruv Raina) is now safe — `new-workspace` is no longer reserved,
so the slug correctly resolves to that workspace and the global route
`/workspaces/new` doesn't shadow it.
2. Refactor workspace_test.go to drive its reserved-slug list from the
reservedSlugs map directly via `for slug := range reservedSlugs`. The
previous hand-copied list was already drifting (40-ish entries vs 58 in
the map). Now drift is impossible.
3. Sweep ~10 stale `/new-workspace` references in code comments to
`/workspaces/new`. Comments only — runtime unchanged. The references
in reserved-slugs.ts/workspace_reserved_slugs.go and CLAUDE.md are
intentionally kept as anti-pattern examples ("don't add hyphenated
word-group root routes like /new-workspace").
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The NoAccessPage button previously only called nav.push('/login'),
leaving the session cookie, React Query cache, and local auth state
intact. AuthInitializer then silently re-authenticates and bounces the
user right back to the workspace URL — the button appeared broken.
Extract the logout flow (clear per-workspace storage, clear cookies,
clear multica_tabs, queryClient.clear(), authStore.logout(), navigate
to /login) into a shared useLogout() hook in packages/views/auth/.
AppSidebar and NoAccessPage both use it now; any future logout entry
point can too.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Desktop tabs persist their full path to localStorage (multica_tabs), so
a tab path like /naiyuan/issues survives app restarts, account switches,
and workspace deletions. Any stale slug caused WorkspaceRouteLayout to
render NoAccessPage immediately on login — the user saw "Workspace not
available" every time they opened the app, with no way to recover
except manually opening a new tab or clearing localStorage.
Root cause: persisted URL strings outlive the server-state they
reference. The auth initializer fetches a fresh workspace list on every
startup, but nothing validated the tab paths against it.
Fix: add tab-store.validateWorkspaceSlugs(validSlugs). Runs on every
change to the workspace list query data (login, background refetch,
realtime workspace:deleted). Any tab whose first path segment isn't in
the valid slug set is reset to `/`, where IndexRedirect picks a live
workspace (or /new-workspace if the user has none). Idempotent, so
over-triggering is safe. Tabs on global paths (/login, /new-workspace,
/invite/...) are left alone.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(daemon): allow startup with zero workspaces
The daemon used to fail fast with "no runtimes registered" when the
initial workspace sync returned zero workspaces. This masked a latent
bug: a newly-signed-up user has no workspaces yet, so the daemon would
crash immediately after login instead of waiting for the first
workspace to be created.
workspaceSyncLoop already polls every 30s (daemon.go:107, 365) to
discover new workspaces — the fail-fast check at startup was bypassing
this dynamic discovery. Remove the check so the daemon stays resident
and picks up the first workspace whenever it appears.
PR #1001 partially addressed this for the "server has workspaces but
local CLI config is empty" case. This finishes the job for the true
zero-workspace state, which until now was masked by the onboarding
wizard always creating a workspace before the daemon started.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor(views): extract CreateWorkspaceForm for reuse
Modal and the upcoming /new-workspace page share the same form +
mutation + slug validation. Extract to a shared component so they
can't drift.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(views): add NoAccessPage for unknown or inaccessible workspace slugs
Rendered when the URL slug doesn't resolve to a workspace the user has
access to. Deliberately doesn't distinguish 404 vs 403 to avoid letting
attackers enumerate workspace slugs.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(paths): add /new-workspace route and reserve slug on both sides
Adds paths.newWorkspace() builder, registers /new-workspace as a global
(pre-workspace) prefix, and reserves the "new-workspace" slug on both
frontend and backend (kept in sync per convention). Existing
"onboarding" reservation retained — removing it would desync FE/BE
and leaves no future fallback if an onboarding route is revived.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* chore(migrations): audit no existing workspace uses 'new-workspace' slug
Migration 046 blocks deploy if any workspace in the DB has slug =
'new-workspace', which would shadow the new global workspace creation
route at /new-workspace.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat: add /new-workspace route on web and desktop
Renders the CreateWorkspaceForm as a full-page workspace creation flow,
used as the destination for first-time users with zero workspaces.
Replaces the 4-step onboarding wizard with a single form.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat: show NoAccessPage on unknown workspace slug, hold null during active removal
Layouts render NoAccessPage when the URL slug doesn't resolve to an
accessible workspace — except when the slug previously resolved during
this layout instance's lifetime.
URL and cache are two asynchronous signals: there will always be a
short window where the URL still points at the old workspace but the
cache has already been invalidated (e.g. just after a delete/leave
mutation, or a realtime workspace:deleted event). Rendering
NoAccessPage during that window would flash "Workspace not available"
with recovery buttons in front of a user who just deleted the
workspace themselves — jarring and wrong.
useWorkspaceSeen classifies the two cases:
- slug was seen before, now gone → user's intent is changing (caller
is navigating away); render null, no flash
- slug never seen → user is genuinely looking at an inaccessible
workspace (stale bookmark, revoked access, link from a former
teammate); render NoAccessPage with recovery options
NoAccessPage deliberately does not distinguish 404 vs 403 to avoid
letting attackers enumerate workspace slugs.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor: redirect zero-workspace users to /new-workspace instead of /onboarding
Switches 8 call sites and the CLI:
- Web: login, auth callback, landing redirect-if-authenticated
- Desktop: routes.tsx IndexRedirect
- Shared: dashboard guard, invite page fallback, workspace-tab on delete,
realtime sync on workspace loss
- CLI: cmd_login.go waitForOnboarding now opens /new-workspace
Also adds /new-workspace to navigation store's lastPath exclusion list
so it doesn't get persisted as a 'last visited' page.
Adds a desktop App.tsx effect that restarts the daemon when workspace
count transitions 0 → ≥1, so first-workspace creation triggers
immediate daemon pickup rather than waiting up to 30s for the daemon's
workspaceSyncLoop.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor: remove onboarding flow
The 4-step onboarding wizard (workspace → runtime → agent → demo issues)
is replaced by:
- /new-workspace: a single-page workspace creation form (Phase 3)
- NoAccessPage: explicit feedback when a slug doesn't resolve (Phase 4)
- daemon zero-workspace bootstrap (Phase 1) so the daemon doesn't
crash before the user creates their first workspace
- desktop daemon restart on first workspace creation (Phase 5) for
instant pickup instead of the 30s workspaceSyncLoop tick
Deletions:
- packages/views/onboarding/ (OnboardingWizard + 4 step components + tests)
- apps/web/app/(auth)/onboarding/page.tsx
- apps/desktop/src/renderer/src/components/onboarding-gate.tsx (+test)
- OnboardingGate wrapper in desktop-layout.tsx
- OnboardingRoute + /onboarding route in desktop routes.tsx
- paths.onboarding() builder + /onboarding from GLOBAL_PREFIXES
- packages/views/package.json onboarding export
- /onboarding from navigation store's EXCLUDED_PREFIXES
Retained (intentional):
- 'onboarding' in RESERVED_SLUGS (both FE + BE) — kept for FE/BE sync
and future-proofing if /onboarding is ever revived
Also drops 4 demo issues that onboarding used to create on the new
workspace ('Say hello', 'Set up repo', etc.). New workspaces are now
fully empty; all list views already render empty-state UI correctly.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* chore: clean stale 'onboarding' references in comments and CLI helpers
Batch cleanup of references to the removed onboarding flow:
- 13 comment sites mentioning 'onboarding' updated to reflect the
new /new-workspace flow or removed where no longer accurate
- CLI waitForOnboarding renamed to waitForWorkspaceCreation (function
name + docstring); behavior unchanged
The 'onboarding' reserved slug entries (frontend + backend) are
intentionally retained — see prior commit rationale.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor(views): extract shared NewWorkspacePage shell
The web (/new-workspace) and desktop (NewWorkspaceRoute) pages had
identical outer layout — same container, heading, and copy — with only
the onSuccess navigation primitive differing. That's exactly the
No-Duplication Rule pattern: extract the shared UI, inject the
platform-specific behavior.
The apps now only own the thin auth guard (web needs it, desktop
routes below WorkspaceRouteLayout already handle it) and the
onSuccess → navigate call.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor: remove rollback compat layer and tighten daemon restart trigger
Two cleanup items:
1. Drop localStorage['multica_workspace_id'] double-write in both
workspace layouts. That write was added as a rollback safety net
for the workspace-slug URL refactor (PR #1138) — the refactor has
since landed and stabilized, so the compat shim is no longer
needed. Per CLAUDE.md: don't keep compat layers beyond their
purpose.
2. Tighten the desktop daemon-restart trigger. The previous ref-based
logic fired a restart on any 0→1 workspace-count transition,
including account switches (user A logout → user B login). Scope
it precisely to 'this session started with zero workspaces and
just gained one' using a three-state ref (null=undecided,
true=empty-start, false=already-restarted-or-started-nonempty).
Account switches are already handled by daemon-manager.ts on
token change, so this avoids a redundant restart there.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(auth): redirect to /login on logout and unauthenticated workspace visits
Two gaps previously left users stuck on blank workspace pages:
1. app-sidebar logout() cleared all state but never moved the URL. The
current path is /{workspaceSlug}/... which has no meaning without
auth; the workspace layout would then see user=null, render null
(via the hasBeenSeen short-circuit), and the user saw a blank page
thinking logout didn't work.
2. The workspace layouts (web + desktop) had no !user handling at all.
Any path that leaves user=null — token expiration, cross-tab logout,
or fresh visit to a workspace URL without a session — resulted in
the same blank screen.
Fix:
- app-sidebar.logout() explicitly push(paths.login()) after authLogout()
to cover the primary (user-initiated) logout path.
- Both workspace layouts get a defensive useEffect that redirects to
/login whenever auth has settled and user is null. Covers token
expiration, realtime logout, and any other silent session loss.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(editor): include done issues in @ mention search
The mention picker filtered against the cached issue list, which only
holds the first page of done issues. Older done issues were unfindable
via @, so users had to hand-write `[MUL-xxx](mention://issue/...)` to
reference them.
Switch the issue portion of the picker to the server-side search
endpoint with `include_closed=true` (matching the global Cmd+K search),
debounced and abortable. Done/cancelled rows render dimmed with a
strikethrough title so they remain visually distinct but selectable.
* fix(editor): unblock member/agent results in @ mention picker
The previous patch made items() async and awaited the server-side issue
search before returning anything, which forced even local member/agent
matches to wait for the 150ms debounce + roundtrip.
Return sync items (members, agents, cached issues) immediately and let
the renderer be updated in-place when extra server results arrive. Also
move the search seq/abort state into the createMentionSuggestion closure
so concurrent ContentEditor instances no longer abort each other's
fetches, and aborts on cleanup so a late response can't write to a
destroyed renderer.
Adds a focused test that locks in the sync member/agent path and the
include_closed=true flag.
GetWorkspaceUsageByDay and GetWorkspaceUsageSummary had the same date
attribution bug as the runtime endpoint fixed in #1167: they bucketed
and filtered on agent_task_queue.created_at (enqueue time), so a task
that queued at 23:58 and reported usage at 00:05 was attributed to the
prior day, and ?days=N became a rolling now()-N window that clipped the
morning of the earliest day returned.
Switch both queries to task_usage.created_at (~= task completion time)
and snap the since cutoff to start-of-day via DATE_TRUNC, mirroring
ListRuntimeUsage.
These endpoints have no frontend caller today, but per offline
discussion they will back the upcoming workspace-level usage dashboard.
Fix preemptively so the dashboard inherits correct numbers.
Add a regression test covering both endpoints with the same
cross-midnight + earliest-day-cutoff scenarios used for runtime usage.
* refactor(runtime): derive runtime usage from task_usage only
The daemon used to scan each runtime's local CLI log directory every 5
minutes (Claude Code, Codex, OpenCode, OpenClaw, Hermes) and post daily
aggregates to /api/daemon/runtimes/{id}/usage. Those directories are
shared with the user's own local CLI sessions, so the user's personal
usage was being counted as Daemon-executed usage. Cursor and Gemini had
no scanner at all, so their runtime-level aggregates were always zero.
Switch GetRuntimeUsage to aggregate task_usage (already scoped to
Daemon-executed tasks) via agent_task_queue.runtime_id. Single source of
truth; Cursor/Gemini/Copilot get runtime usage for free; no reliance on
external CLI log formats.
Removes:
- server/internal/daemon/usage/ (all scanners)
- Daemon.usageScanLoop + providerToRuntimeMap
- Client.ReportUsage
- ReportRuntimeUsage handler + POST /api/daemon/runtimes/{id}/usage
- UpsertRuntimeUsage / GetRuntimeUsageSummary queries
- runtime_usage table (migration 046)
Refs: MUL-786
* fix(runtime): bucket daily usage by task_usage.created_at, not enqueue time
ListRuntimeUsage was aggregating by DATE(atq.created_at) and filtering
on atq.created_at. agent_task_queue.created_at is the enqueue timestamp,
which drifts from actual token-production time: a task queued at 23:58
and executed at 00:05 was attributed to yesterday; a task sitting in
the queue overnight was counted on the queue day.
The ?days=N cutoff also became a rolling window (now() - N) instead of
a calendar-day boundary, silently clipping the morning of the earliest
day returned.
Switch bucket + filter to task_usage.created_at (~= task completion /
usage-report time) and snap the since cutoff to start-of-day via
DATE_TRUNC.
Add a regression test covering both scenarios: cross-midnight task
attributes to the day tokens were reported, and the earliest day's
pre-cutoff rows are still included.
The Run History list only had the 'Issue linked' text as a click target.
Wrap the entire row in AppLink when an issue is linked so the whole row
navigates to the issue.
Every other backend (Claude, Gemini, OpenCode, OpenClaw, Hermes) honors
ExecOptions.ResumeSessionID — only Codex didn't. That's why users on
the Codex runtime saw each new comment on an issue start a fresh Codex
conversation: the daemon persists Result.SessionID per (agent, issue)
and passes it back as PriorSessionID, but codex.go always called
thread/start and never populated SessionID, so the value round-tripped
as empty.
Wire the missing half:
- Extract startOrResumeThread on codexClient. When ResumeSessionID is
set, call thread/resume (per the Codex app-server protocol), passing
only cwd / model / developerInstructions overrides so the thread
keeps its persisted model and reasoning effort. If resume fails
(unknown thread, schema drift, transport error) fall back to
thread/start so the task still runs on a fresh thread.
- Surface the live threadID as Result.SessionID on the final emit so
the daemon stores it and feeds it back into ResumeSessionID on the
next claim.
Tests drive the new helper through the fake stdin harness, covering:
fresh start, successful resume, fallback on resume error, fallback
when resume returns no thread ID, and surfacing of thread/start
failures.
AuthStore.initialize() cleared the stored token on any error from
`api.getMe()`, which meant a transient failure — backend rolling
restart, network blip, HMR-aborted fetch in local dev — would force
a re-login. On 401 the token is already cleared upstream via
ApiClient.onUnauthorized, so the store's catch block only needs to
reset the in-memory user state.
Check `err instanceof ApiError && err.status === 401` before clearing
workspace context; leave the token in storage for every other error
so the next initialize() can retry.
Adds regression tests covering the 401 / 500 / network-failure / happy
paths.
Problem
-------
The v2 workspace URL refactor (#1141) switched the frontend from sending
X-Workspace-ID (UUID) to X-Workspace-Slug. The workspace middleware was
updated to accept the slug and translate it via GetWorkspaceBySlug.
But the handler package maintained a PARALLEL resolver
(`resolveWorkspaceID` in handler.go) used by endpoints that sit outside
the workspace middleware — and that resolver was never updated. It only
checked context / ?workspace_id / X-Workspace-ID, never the slug.
/api/upload-file is the one production route that hit the broken path:
it's user-scoped (not behind workspace middleware) because it also
serves avatar uploads (no workspace). Post-refactor requests from the
frontend arrived with only X-Workspace-Slug; the handler resolver
returned "", the code fell into the "no workspace context" branch, and
every file upload since v2 landed in S3 with no corresponding DB
attachment row — files orphaned, invisible to the UI.
Root cause is structural: two resolvers doing the same job, written
independently, diverged silently when one was updated.
Fix
---
Collapse to a single shared helper. middleware.ResolveWorkspaceIDFromRequest
is the new canonical resolver; both the middleware's internal
`resolveWorkspaceUUID` (for middleware gating) and the handler-side
`(h *Handler).resolveWorkspaceID` (promoted from a package function)
now delegate to it. Priority order matches what the middleware has had
since v2: context > X-Workspace-Slug header > ?workspace_slug query >
X-Workspace-ID header > ?workspace_id query.
Impact analysis
---------------
47 call sites of the old `resolveWorkspaceID(r)` are renamed to
`h.resolveWorkspaceID(r)`. 46 of them sit behind workspace middleware,
so they hit the context fast path and see zero behavior change. The
one caller that actually gains capability is UploadFile — which now
correctly recognizes slug requests and creates DB attachment rows.
Tests
-----
- New table-driven unit test for ResolveWorkspaceIDFromRequest covers
all priority levels and the unknown-slug fallback.
- Regression tests for UploadFile: once with X-Workspace-Slug only
(the broken path), once with X-Workspace-ID only (legacy CLI/daemon
compat path). Both assert that a DB attachment row is created.
- Full Go test suite passes; typecheck + pnpm test unaffected.
Plan
----
See docs/plans/2026-04-16-unify-workspace-identity-resolver.md for the
full first-principles writeup.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Problem
-------
On desktop, creating a new tab triggered thousands of chat-store
rehydration logs per second (sustained for seconds). Same session,
same workspace — nothing actually changed. `pnpm test` was clean; the
bug only manifests at runtime with React 19 Activity + multi-tab.
Root cause
----------
Every tab's WorkspaceRouteLayout kept its own `syncedSlugRef` to decide
"did slug change since last sync". That model assumes one layout
instance equals one workspace context — true on web, false on desktop
where N tabs each mount their own layout. Activity remounts +
tab-router-sync stirring the tab store caused per-layout refs to drift
out of agreement with the module-level truth, so each ref independently
called `rehydrateAllWorkspaceStores()`. The existing microtask dedup
only coalesced same-tick calls; successive ticks each scheduled another
iteration through every registered rehydrate fn.
Fix
---
Move the "did slug actually change?" decision to where the truth lives:
inside `setCurrentWorkspace` itself. The singleton now:
- Returns immediately when the slug is already current (idempotent).
- Fires slug subscribers + persist rehydrate as internal side effects
when (and only when) the slug transitions.
Layouts are simplified to "feed the URL slug in"; they no longer
maintain a ref guard or call rehydrate explicitly. N tabs feeding the
same slug is naturally a no-op after the first — the model no longer
depends on "one layout instance" as an implicit invariant.
Also hardens the original render-time race that motivated the v2
refactor: both layouts now gate on `!listFetched || !workspace` so
`useWorkspaceId()` in descendants is guaranteed non-null.
Public API
----------
`rehydrateAllWorkspaceStores` removed from `@multica/core/platform`
exports — it's now purely an internal effect of `setCurrentWorkspace`.
The function itself is deleted; the rehydrate loop lives inline in
`setCurrentWorkspace`.
Tests
-----
Four new tests covering the new semantics: single rehydrate on mount,
same-slug noop across repeat calls, real workspace switch fires again,
logout → re-entry into same workspace fires again.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add a "Download Desktop" button in the hero section alongside the
existing CTA and GitHub buttons, linking to the latest GitHub release.
Also add a Desktop link in the footer product group for both EN and ZH.
* feat(agent): add GitHub Copilot CLI backend
Integrate Copilot CLI as a new agent backend using the stable
`-p` JSONL mode (`--output-format json`), following the same
spawn-CLI-scan-JSONL pattern established by claude.go.
Backend (server/pkg/agent/copilot.go):
- Spawn `copilot -p <prompt> --output-format json --allow-all-tools --no-ask-user`
- Parse streaming JSONL events (system/assistant/user/result/log)
- Extract session ID for resume support (`--resume <id>`)
- Accumulate per-model token usage for billing
- Filter blocked args to prevent protocol-critical flag overrides
Daemon config:
- Probe MULTICA_COPILOT_PATH / MULTICA_COPILOT_MODEL env vars
- Copilot uses AGENTS.md (native discovery) and default skills path
Frontend:
- Add Copilot logo SVG and provider switch case
Tests: 14 unit tests covering arg building, event parsing, usage
accumulation, and edge cases. All Go + TS checks pass.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(daemon): add restart subcommand, make daemon uses it
- `daemon start` keeps original behavior: errors if already running
- `daemon restart` stops existing daemon then starts fresh
- `make daemon` now runs `daemon restart --profile local`
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(copilot): address review nits 1-5
- Nit 1: Add MinVersions["copilot"] = "1.0.0"
- Nit 2: Seed activeModel from session.start.data.selectedModel (falls
back to opts.Model, then "copilot"). First-turn tokens now get correct
model attribution.
- Nit 3: Handle assistant.reasoning/reasoning_delta → MessageThinking,
reasoningText in assistant.message → MessageThinking,
session.warning → MessageLog{warn}
- Nit 4: Extract handleCopilotEvent() method shared by production and
tests — no more duplicated switch body that can drift
- Nit 5: Deltas write to output buffer as defense-in-depth; if process
dies before assistant.message, output is non-empty
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
When codex emits `turn/completed` with `status="failed"` or a terminal
top-level `error` notification, the daemon previously treated the turn
as successfully completed, saw no accumulated text, and surfaced the
generic "codex returned empty output" — hiding the real reason (auth,
sandbox, API error, etc.).
Capture `turn.error.message` on failed turns and the `error.message`
from non-retrying top-level error notifications, then propagate them
through `Result.Error` with `finalStatus="failed"` so the daemon's
default branch reports the actual cause.
Dev mode now uses a separate app name ('Multica Dev') and userData path
before acquiring the single-instance lock, so the lock file no longer
collides with the packaged production app. The AppUserModelId is also
differentiated (ai.multica.desktop.dev vs ai.multica.desktop).
This follows the same pattern VS Code uses for Stable / Insiders
coexistence: isolate identity before requestSingleInstanceLock().
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* docs: add Pi and Gemini runtimes to supported-agent references
CLI_AND_DAEMON.md, SELF_HOSTING.md, and SELF_HOSTING_ADVANCED.md listed
claude/codex/opencode/openclaw/hermes as supported runtimes in their agent
tables and env-var overrides but omitted the pi and gemini entries that
the daemon already registers (server/internal/daemon/config.go).
* docs(readme): list all supported runtimes (add Hermes, Gemini, Pi)
* docs: add Cursor runtime, fix Pi URL, clarify daemon ASCII diagram
- Add Cursor Agent (cursor-agent CLI, MULTICA_CURSOR_PATH/MODEL) to the
supported-runtime tables, env-var lists, and prose across README,
CLI_AND_DAEMON, CLI_INSTALL, SELF_HOSTING, and SELF_HOSTING_ADVANCED.
- Fix Pi's canonical URL from github.com/paperclipai/paperclip to
https://pi.dev/.
- Rework the Agent Daemon box in both READMEs so provider names live in
an annotation outside the box instead of being wrapped mid-word
(`OpenClaw/Code`), which read as a phantom "Code" runtime.
* feat(agent): add Cursor Agent CLI runtime support
Add cursor-agent as a new agent backend, following the same pattern as
existing providers. The implementation spawns cursor-agent CLI with
stream-json output, parses JSONL events into the unified Message type,
and supports session resume, usage tracking, and auto-approval (--yolo).
Changes:
- server/pkg/agent/cursor.go: cursorBackend implementation
- server/pkg/agent/cursor_test.go: unit tests for args, parsing, errors
- server/pkg/agent/agent.go: register "cursor" in New() factory
- server/internal/daemon/config.go: probe cursor-agent in PATH
- server/internal/daemon/execenv/context.go: cursor skill discovery path
- server/internal/daemon/execenv/runtime_config.go: AGENTS.md injection
- packages/views/.../provider-logo.tsx: cursor logo in UI
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(agent): address PR review for cursor backend
1. Fix token usage double-counting: usage is now taken exclusively from
"result" events (session totals). Per-message usage in "assistant"
events is intentionally ignored. "step_finish" usage is only used as
fallback when no "result" usage is available.
2. Remove dead code: isCursorUnknownSessionError() and its regex were
defined but never called. Removed along with corresponding test.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(agent): add missing CustomArgs, SystemPrompt, MaxTurns, and debug logging to cursor backend
- Add cursorBlockedArgs and filterCustomArgs support for safe custom arg passthrough
- Add --system-prompt and --max-turns flag support to buildCursorArgs
- Add debug logging of command args before execution (consistent with all other backends)
- Move stdout-close goroutine inside main goroutine (consistent with claude.go pattern)
- Add tests for SystemPrompt/MaxTurns and CustomArgs filtering
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* chore: make daemon uses local profile & update Cursor logo to official brand
- Makefile: make daemon now runs 'daemon start --profile local' for local dev
- Replace Cursor runtime logo with official brand SVG (removed background rect)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(agent): remove unsupported --system-prompt and --max-turns from cursor-agent
cursor-agent CLI does not support these flags. Instructions are already
injected via AGENTS.md and .cursor/skills/ files.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* fix(agent): prevent step_finish + result usage double-counting in cursor
Split usage accumulation into separate stepUsage and resultUsage maps.
After stream ends, use resultUsage if available (session totals from
result event), otherwise fall back to stepUsage (sum of step_finish).
This prevents 2x counting when result.usage already includes totals.
Added table-driven test covering: result-only, step_finish-only,
step_finish+result (no double count), and multi-model scenarios.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
* docs(agent): fix misleading comment on cursor -p flag
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---------
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: yushen <ldnvnbl@gmail.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Replace the placeholder Greek-letter π glyph with pi.dev's actual
pixel-art "pi" wordmark on the brand's dark background. Source:
https://pi.dev/logo.svg.
* feat(agent): add Pi agent runtime support
Add Pi as a new agent runtime provider, following the established adapter
pattern. Pi CLI outputs JSONL events which are parsed for messages, tool
calls, and usage tracking.
Backend:
- New piBackend implementing the Backend interface (pi.go)
- Pi CLI discovery via MULTICA_PI_PATH env var or PATH lookup
- JSONL event stream parsing (agent_start, message_update, thinking_update,
tool_execution_start/end, agent_end)
- Usage scanner for ~/.pi/sessions/*.jsonl files
- Runtime config injection via AGENTS.md
- Skill injection to .pi/agent/skills/
Frontend:
- Pi provider logo (teal π icon)
- Pi label in transcript dialog
Docs:
- Updated all provider lists in README, CLI_INSTALL, and docs
* fix(agent): filter Pi usage scanner to agent_end events only
Address review feedback: restrict usage parsing to agent_end events
which contain cumulative totals, preventing potential inaccuracy if
Pi adds usage fields to other event types in the future.
* fix(agent): align Pi runtime with real CLI flags, event schema, and custom_args
- Flags: Pi's CLI uses `--mode json` (not `--output-format jsonl`), has no
`--yolo` (explicit `--tools` allowlist instead), takes the prompt as a
positional argument (not `-p <prompt>`), splits model as
`--provider <name> --model <id>`, and treats `--session` as a file path
that must exist before spawn.
- Event parsing: rewrite the stream event struct to match Pi's actual
JSON event schema (`message_update.assistantMessageEvent.delta`,
`turn_end.message.usage.{input,output,cacheRead,cacheWrite}`, etc.).
- Sessions: generate/persist session files under ~/.multica/pi-sessions/
and use the file path as the opaque SessionID returned to the daemon.
- Usage scanner: read assistant `message` events from the same session
files (Pi's session-file schema, distinct from the stdout stream).
- Custom args: consume `ExecOptions.CustomArgs` via `filterCustomArgs`
with a Pi-specific blocked set (`-p`, `--print`, `--mode`, `--session`)
so Pi matches the pattern shared by every other agent backend.
GetActiveTaskForIssue, CancelTask, ListTasksByIssue, and GetIssueUsage
accepted the issueId URL parameter and queried by it without verifying
that the issue belonged to the caller's X-Workspace-ID workspace. The
RequireWorkspaceMember middleware only proves membership in the header
workspace; it does not bind the path-parameter issue to it. A member of
workspace A could therefore enumerate tasks, cancel tasks, and read
usage metadata for any issue UUID in workspace B.
Route every issueId through loadIssueForUser (matching GetIssue and the
existing comment/subscriber handlers). For CancelTask additionally
verify that the task's IssueID matches the loaded issue — the task must
not only belong to the caller's workspace but also to the specific
issue named in the URL, and the access check must run before any
mutation.
Follow-up to MUL-899 / #1112.
The top bar pads `pl-20` for the macOS traffic lights only when
`state === "collapsed"`, but the shadcn sidebar also hides itself in
mobile mode (<768px) where `state` stays `"expanded"` and only
`isMobile` flips. In that case tabs slid under the traffic lights and
no UI affordance existed to bring the sidebar back (since the in-sidebar
trigger went off-canvas with it).
Treat both as "sidebar not in main flow", apply the padding, and render
a `SidebarTrigger` in the header (with `no-drag` so the window drag
region doesn't swallow the click).
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Every other backend (claude, codex, opencode, openclaw, gemini) filters
opts.CustomArgs through a per-backend blocked map so protocol-critical
flags can't be overridden via the Create Agent UI. The hermes backend
appended CustomArgs directly to argv, so any future flag we add to the
map would be silently bypassed here.
Add hermesBlockedArgs (with 'acp' as the pinned subcommand) and route
CustomArgs through filterCustomArgs. Behaviour is identical for today's
use cases; the change prevents accidental protocol-flag overrides and
brings hermes in line with the other five backends.
Closes#1113
Co-authored-by: shaun0927 <shaun0927@users.noreply.github.com>
Adds TestGetIssueGCCheck_WithDaemonToken_CrossWorkspace alongside the
existing TestGetTaskStatus_WithDaemonToken_CrossWorkspace, covering:
- daemon token scoped to a different workspace → 404 (matches the
"issue not found" status, so no UUID enumeration oracle)
- daemon token scoped to the issue's workspace → 200 with status and
updated_at fields populated
Follow-up to #1121, which fixed the underlying IDOR reported in #1112
but did not ship a regression test. This gates the class of bug at CI
so the next handler to forget requireDaemonWorkspaceAccess will be
caught before merge.
After the slug-first URL refactor, the frontend sends X-Workspace-Slug
and the workspace middleware resolves it into a UUID stored in the
request context. The inbox handlers still read X-Workspace-ID directly
from the request header, which is now absent, so every inbox query ran
with an empty workspace_id and returned zero rows.
Switch all six inbox handlers to ctxWorkspaceID(r.Context()), matching
the pattern already used by chat / issue / project / autopilot. No
frontend changes required — the slug header path was already correct.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The daemon GC check endpoint did not verify the caller's access to the
issue's workspace, letting a daemon token or PAT scoped to workspace A
read issue status/updated_at for any issue UUID across the instance.
Mirror the pattern used by every other handler in daemon.go: look up
the issue's workspace and gate on requireDaemonWorkspaceAccess.
Closes#1112
Co-authored-by: shaun0927 <shaun0927@users.noreply.github.com>
Follow-up to #1126 (which closed the HTML-injection vector in the Body).
The Subject line is not HTML-rendered, so html.EscapeString would leak
literal entities into recipient inboxes. Instead:
- Strip control characters from workspace/inviter names (defense in depth
even though Resend also filters CR/LF).
- Cap each field at 60 runes so an attacker can't stuff a full phishing
pitch into a workspace name that gets sent from noreply@multica.ai.
Also extracts buildInvitationParams to make the sanitization logic
testable without mocking the Resend SDK, and adds a test covering:
- HTML escape behavior for script/attribute/anchor injection payloads
- Subject stripping of \r\n\t and other unicode controls
- Subject NOT being HTML-escaped (so "Acme & Co." stays literal)
- Subject length bounds
- Benign inputs pass through unchanged
Adds a note on SendVerificationCode that its body uses only
server-generated content, to prevent the same pitfall from creeping in.
Refs #1117
* fix(email): HTML-escape workspace/inviter names in invitation email
SendInvitationEmail interpolated workspaceName and inviterName directly
into the HTML body via fmt.Sprintf with no escaping. A workspace owner
who sets a name like '</h2><a href="https://evil.example">Click</a>'
can break the email structure and inject attacker-controlled links that
appear as part of the official Multica invitation.
Escape both values with html.EscapeString before interpolation. The
Subject line also gets the escaped variants since some transports render
HTML-entity-like sequences.
Closes#1117
* fix(email): use raw names in Subject, keep HTML-escape for body only
Email Subject is a plain-text context — applying html.EscapeString
turns "A&B" into "A&B" and "O'Brien" into "O'Brien" in the
recipient's inbox. Keep the escape for the Html body where it prevents
injection, but use the original values in Subject.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: shaun0927 <shaun0927@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* Reapply "feat: workspace URL refactor + slug-first API identity (#1131)" (#1137)
This reverts commit 9b94914bc8.
* compat: legacy URL redirect + localStorage double-write for safe rollback
The first attempt at this refactor (#1131) was reverted because existing
users on old URLs (/issues, /projects, etc.) hit 404 immediately after
deploy, and rolling back left them with empty dashboards — the legacy
code reads localStorage["multica_workspace_id"] to attach a workspace
to API requests, but the new code had stopped writing that key.
Two compat layers added on top of the restored refactor:
1. proxy.ts now intercepts legacy route prefixes (/issues/*, /projects/*,
/agents/*, /inbox/*, /my-issues/*, /autopilots/*, /runtimes/*,
/skills/*, /settings/*). Logged-in users with a last_workspace_slug
cookie are 302'd to /{slug}/{rest}, preserving their deep link. Users
without the cookie bounce through / where the landing page picks a
workspace client-side. Unauthenticated users go to /login.
2. Both layouts now double-write the workspace id to the legacy
localStorage key on every workspace entry. New code ignores this key
— it exists solely so that if this PR ever gets reverted again, the
legacy build reading the key would still find the correct workspace
and avoid the empty-dashboard symptom users saw during the rollback.
Net effect: any direction of deploy ↔ rollback is now cache-compatible,
and any direction of old bookmark → new route resolves without 404.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(platform): defer rehydrateAllWorkspaceStores to a microtask
Same React 19 render-phase restriction that forced setCurrentWorkspace
to defer its subscriber notifications. rehydrateAllWorkspaceStores
synchronously calls each persist store's rehydrate, which setState()s
the store, which schedules updates on any subscribed component. When
the workspace layout's render-phase ref guard invoked this, React
complained that SearchCommand (a store subscriber) couldn't be
re-rendered while WorkspaceLayout was still rendering.
Fix: queueMicrotask the rehydrate loop and add a pending-flag guard so
rapid workspace switches coalesce into one rehydrate on the final slug.
Persist stores tolerate one microtask of staleness — they hold UI
preferences, not correctness-critical state.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat: workspace URL refactor + slug-first API identity
Make the URL the single source of truth for workspace identity.
All workspace-scoped URLs now carry the workspace slug as the first
path segment (/{slug}/issues, /{slug}/projects, etc.), matching the
industry standard (Linear, Notion, Vercel, GitHub).
## Key architectural changes
**URL-driven workspace identity:**
- Web routes moved under app/[workspaceSlug]/(dashboard)/
- Desktop routes nested under /:workspaceSlug
- paths.ts builder centralises all URL construction
- reserved-slugs validation (backend + frontend + DB migration audit)
**Slug-first API contract:**
- Frontend sends X-Workspace-Slug header (from URL) instead of X-Workspace-ID (UUID)
- Backend middleware resolves slug → UUID via GetWorkspaceBySlug, falls back to
X-Workspace-ID for CLI/daemon backwards compatibility
- WebSocket auth accepts ?workspace_slug query param with SlugResolver callback
**State cleanup:**
- Deleted: useWorkspaceStore (Zustand mirror), switchWorkspace/hydrateWorkspace/
clearWorkspace, localStorage["multica_workspace_id"], api._workspaceId
- useCurrentWorkspace() derives from URL slug + React Query workspace list
- useWorkspaceId() is now a bridge hook (no Context, derives from useCurrentWorkspace)
- WorkspaceIdProvider removed from DashboardGuard
- Paired module vars (slug + UUID) in workspace-storage.ts for non-React consumers
**Layout simplified:**
- Render-phase ref guard sets workspace context synchronously (no async gate)
- DashboardGuard handles auth redirect, loading state, and workspace resolution
- Subscriber notifications deferred via queueMicrotask (React 19 compat)
- persist namespace uses slug (immutable) instead of UUID
## Issues resolved
MUL-43 (share links), MUL-509 (mobile workspace switch), MUL-723 (workspace in URL),
MUL-727 (create workspace flash), MUL-728 (delete workspace no-navigate),
MUL-820 (sidebar Join not switching)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: resolve code review C3/C4/C5/C6 — desktop deadlock + hardcoded paths
C3: Desktop OnboardingGate was calling useCurrentWorkspace() outside
WorkspaceSlugProvider → always null → permanent onboarding deadlock.
Rewrite to use useQuery(workspaceListOptions()) which reads React Query
cache directly without slug context. Remove DashboardGuard from
DesktopShell (auth gating handled by AppContent, workspace routing by
WorkspaceRouteLayout per-tab).
C4: Landing page "Dashboard" links hardcoded /issues (no longer valid).
Changed to / — proxy handles redirect to /{lastSlug}/issues.
C5: autopilots-page.tsx had one hardcoded /autopilots/${id} link.
Changed to wsPaths.autopilotDetail(id).
C6: inbox-page.tsx hardcoded /inbox paths. Changed to wsPaths.inbox().
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(desktop): wrap shell in WorkspaceSlugProvider from module var
AppSidebar calls useWorkspacePaths() → useRequiredWorkspaceSlug() which
throws outside WorkspaceSlugProvider. In the desktop shell, the sidebar
renders at the shell level (outside any tab's WorkspaceRouteLayout).
Fix: DesktopShell reads the current slug via useSyncExternalStore on
the workspace-storage singleton. When slug is available, wraps the
entire shell in WorkspaceSlugProvider. When null (first mount before
any tab's WorkspaceRouteLayout sets it), shows a loading spinner.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(desktop): migrate old tab paths + fix shell slug deadlock
Tab store rehydration: old-format paths like "/issues/abc" (missing
workspace slug prefix) are reset to "/" so IndexRedirect picks the
correct workspace. Detection: if the first segment is a known route
name (issues, projects, etc.) rather than a workspace slug, it's an
old-format path.
Desktop shell: TabContent must always render (not gated behind slug
check) so WorkspaceRouteLayout can mount and call setCurrentWorkspace.
Only sidebar and shell-level UI (chat, modals, search) gate on slug
being present.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two editor bugs fixed:
1. Descriptions saved unnecessarily on every document change (no dirty
check). Added onCreate baseline capture + string comparison in the
debounced onUpdate handler so mutations only fire when content
actually changes.
2. Clearing a description didn't persist — empty string was converted
to undefined via `md || undefined`, causing the field to be omitted
from the API request. Changed to `md` so empty strings reach the
backend and clear the description via COALESCE.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- TrimSpace incoming repoURL in ensureRepoReady to prevent unnecessary
server refreshes when CLI passes URLs with whitespace
- Add comment on reposVersion field clarifying it is stored for future
version-based skip optimization
- Add concurrency safety comment on syncWorkspacesFromAPI skip logic
- Add test for URL trimming fast-path behavior
* feat(server): trigger agent when issue moves out of backlog
When a member moves an agent-assigned issue from "backlog" to an active
status (e.g. "todo", "in_progress"), enqueue an agent task so the agent
starts working. This lets backlog act as a parking lot where issues can
be assigned to agents without immediately triggering execution.
Applies to both single and batch issue updates.
* fix(server): treat backlog as parking lot — no trigger on create/assign
Address review feedback: creating or assigning an agent to a backlog
issue no longer triggers immediate execution. Only moving out of backlog
to an active status triggers the agent, producing exactly one task.
- shouldEnqueueAgentTask now gates on backlog status
- backlog→active trigger uses isAgentAssigneeReady directly
- Added TestBacklogNoTriggerOnCreate test
- Updated TestBacklogToTodoTriggersAgent to assert exactly 1 task
across the full create→move path (no manual cleanup)
* feat(ui): show toast hint when assigning agent to backlog issue
Users may not know that backlog issues won't trigger agent execution
until moved to an active status. Show an actionable toast with a
"Move to Todo" button when:
- Assigning an agent to a backlog issue in the detail page
- Creating a backlog issue with an agent assignee
* feat(ui): add "Don't show again" option to backlog agent toast
Users who understand the backlog parking lot behavior can dismiss the
hint permanently. Uses localStorage to persist the preference.
* feat(ui): replace backlog agent toast with AlertDialog
Use a modal dialog instead of a toast notification so users must
explicitly acknowledge the hint. The dialog offers three options:
- "Move to Todo" — changes status and triggers the agent
- "Keep in Backlog" — dismisses without action
- "Don't show again" — persists dismissal in localStorage
* fix(ui): improve backlog agent dialog
* fix(ui): close create dialog behind hint, use checkbox for don't-show-again
1. Create Issue dialog now closes when the backlog agent hint appears,
so only the hint dialog is visible (not stacked behind).
2. "Don't show again" is now a checkbox instead of a separate button.
When checked, clicking either "Keep in Backlog" or "Move to Todo"
persists the preference.
* fix(ui): smooth backlog agent hint dialog
* fix(test): add useUpdateIssue mock to create-issue test
The test mock for @multica/core/issues/mutations was missing the
useUpdateIssue export that create-issue.tsx now imports, causing
CI failure.
DESKTOP_SPAWN_ENV was a top-level const in daemon-manager.ts that
snapshotted process.env at module load. Because ESM imports are hoisted
and evaluated before main/index.ts runs fix-path, the snapshot captured
launchd's minimal PATH — missing ~/.local/bin, Homebrew, etc. The main
process then had the corrected PATH, but every spawned daemon inherited
the stale one and failed with "no agent CLI found" on fresh GUI launches.
Convert it to desktopSpawnEnv() so process.env is read at call time,
after fix-path has already updated it.
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The desktop login was reading VITE_WEB_URL, which is defined nowhere
in the committed env files. In production builds the variable was
undefined, so Google login opened http://localhost:3000/login?platform=desktop
instead of https://multica.ai/login?platform=desktop.
Switch to VITE_APP_URL, which is already set in apps/desktop/.env.production
and is the same variable platform/navigation.tsx uses for shareable links.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Fresh desktop accounts no longer need to walk through runtime, agent,
and get-started steps before reaching the app. Once the workspace is
created, the onboarding gate hands off directly to the main shell.
Web onboarding is unchanged.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(desktop): add macOS app icon
Replace the default electron-vite scaffold icon with the Multica asterisk
icon. Adds build/icon.icns so electron-builder picks it up automatically
via the `buildResources: build` config — no YAML change needed.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(desktop): run electron-vite build inside package script
The package wrapper only ran bundle-cli.mjs and electron-builder, so
electron-builder silently packaged whatever was already in out/. On a
fresh checkout (or after a partial build) this shipped an app with a
missing renderer bundle, which white-screens on launch.
Add an explicit `electron-vite build` step between bundle-cli and
electron-builder so `pnpm package` is self-contained.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(desktop): restore shell PATH in main process for GUI launches
macOS/Linux GUI launches inherit a minimal PATH from launchd that omits
~/.zshrc, Homebrew, nvm, ~/.local/bin, and other shell config. Child
processes spawned from the main process — including the bundled multica
CLI used by daemon-manager — inherit the same stripped PATH, so the CLI
fails to locate agent binaries like claude, codex, opencode, etc. with
"no agent CLI found: … ensure it is on PATH".
Use `fix-path` to recover the real shell PATH at startup, then prepend
common install locations (/opt/homebrew/bin, /usr/local/bin,
~/.local/bin) as a fallback for broken shell rc or non-interactive
$SHELL. Runs before setupDaemonManager so every subsequent spawn sees
the corrected PATH.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(desktop): show onboarding wizard when authed user has no workspace
Desktop is a single-shell architecture — every route, including
/onboarding, lives inside DashboardGuard. The guard returns its loading
fallback whenever workspace is null, so a fresh account that logs in
with no workspaces ends up stuck on the spinner forever: the
`replace(onboardingPath)` redirect navigates the tab router, but
DashboardGuard still blocks its children because workspace is still
null.
Handle the empty-workspace case in DesktopShell itself: render
OnboardingWizard as a full-screen takeover, bypassing DashboardGuard.
A ref-based flag freezes the "needs onboarding" decision at first
mount so creating a workspace mid-wizard (step 0) doesn't unmount the
wizard and dump the user into the main shell before steps 1-3
(runtime, agent, get started) finish.
Also add a local `bootstrapping` flag in AppContent so DesktopShell
doesn't mount until the deep-link login chain (loginWithToken →
syncToken → listWorkspaces → hydrateWorkspace) fully resolves. Without
it, the shell would briefly see `!workspace` before hydration lands,
causing users with existing workspaces to flash the wizard (or, with
the ref freeze, get stuck in it permanently).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor(desktop): extract OnboardingGate with test coverage
Pull the "render onboarding wizard when authed user has no workspace"
logic out of DesktopShell into a dedicated OnboardingGate component.
Replaces the ref-based freeze with a lazy useState initializer
(`useState(() => !hasWorkspace)`), which is React's idiomatic pattern
for "capture a value once at mount". The freeze semantics are unchanged:
creating a workspace in step 0 of the wizard must not unmount it,
because steps 1-3 still need to run; only `onComplete` flips the gate
back to the main shell.
Also de-duplicates the wrapping DesktopNavigationProvider — both branches
of the shell now share a single provider instead of re-mounting one per
branch.
Wire up jsdom + @testing-library/react in the desktop vitest config
(mirroring packages/views) and add three deterministic tests covering:
1. children render when hasWorkspace is true at mount
2. wizard stays mounted when hasWorkspace flips to true mid-flow
3. onComplete transitions the gate to children
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor(desktop): drop redundant syncToken call in deep-link login
daemonAPI.syncToken was called twice on a deep-link login: once inside
the deep-link handler's bootstrapping chain, and again in the
useEffect([user]) that reacts to the user state change. Both calls spawn
a multica CLI subprocess over IPC, wasting ~1-2s of startup time on the
critical login path.
Keep the [user] effect (it covers the session-restore path too) and
drop the explicit call from the deep-link handler. Net effect: login
latency shrinks, behavior is unchanged.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(selfhost): persist local uploads and proxy file routes
* fix(selfhost): keep local uploads across container recreation
* docs(selfhost): restore relative local upload dir example
Document the complete workflow for running backend, frontend, and daemon
from source in a fully isolated environment. Covers dynamic profile
naming, automated auth, Desktop app testing, and cleanup — all without
touching the system CLI config or production environment.
GoReleaser produces .zip for Windows and .tar.gz for other platforms,
but the update command hardcoded .tar.gz for all platforms, causing a
404 error on Windows.
- Select .zip extension when runtime.GOOS is "windows"
- Add extractBinaryFromZip() for zip archive extraction
- Use "multica.exe" as the binary name on Windows
Closes#1072
Ensures the database schema is always up to date when starting the app,
preventing silent API failures caused by missing columns after pulling
latest changes.
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Add a debug-level log line in every agent backend (claude, codex,
opencode, openclaw, gemini, hermes) that prints the executable path
and full argument list when spawning the agent process. Helps diagnose
custom args, model overrides, and other CLI flag issues.
2026-04-15 16:21:55 +08:00
438 changed files with 33292 additions and 7366 deletions
@@ -133,6 +133,7 @@ make start-worktree # Start using .env.worktree
- Unless the user explicitly asks for backwards compatibility, do **not** add compatibility layers, fallback paths, dual-write logic, legacy adapters, or temporary shims.
- If a flow or API is being replaced and the product is not yet live, prefer removing the old path instead of preserving both old and new behavior.
- Avoid broad refactors unless required by the task.
- New global (pre-workspace) routes MUST use a single word (`/login`, `/inbox`) or a `/{noun}/{verb}` pair (`/workspaces/new`). NEVER add hyphenated word-group root routes (`/new-workspace`, `/create-team`) — they collide with common user workspace names and force endless reserved-slug audits. Reserving the noun (`workspaces`) automatically protects the entire `/workspaces/*` subtree.
### Package Boundary Rules
@@ -161,7 +162,7 @@ When the two apps need different behavior for the same concept (e.g., different
When adding a new page or feature:
1.**New page component** → add to `packages/views/<domain>/`. Never import from `next/*` or `react-router-dom`.
2.**Wire it in both apps** → add a route in `apps/web/app/` (Next.js page file) AND in the desktop router.
2.**Wire it in both apps** → add a route in `apps/web/app/` (Next.js page file) AND in the desktop router.**Exception**: pre-workspace transition flows (create workspace, accept invite) are NOT routes on desktop — they're `WindowOverlay` state. See *Desktop-specific Rules → Route categories*.
3.**Navigation** → use `useNavigation().push()` or `<AppLink>`. Never use framework-specific link/router APIs in shared code.
4.**Shared guards/providers** → use `DashboardGuard` from `packages/views/layout/`. Don't create separate guard logic per app.
5.**Platform-specific UI** → if a feature is web-only or desktop-only, keep it in the respective app. Use props slots (`extra`, `topSlot`) on shared layout components to inject platform-specific UI.
@@ -175,6 +176,70 @@ Both apps share the same CSS foundation from `packages/ui/styles/`.
- **Shared styles** → `packages/ui/styles/`. Never duplicate scrollbar styling, keyframes, or base layer rules in app CSS.
- **`@source` directives** → both apps scan shared packages so Tailwind sees all class names.
## Desktop-specific Rules
These rules apply to `apps/desktop/` only. Web has different constraints (URL bar, SSR, no tabs) and doesn't share these concerns. Every rule in this section was added after a concrete bug — treat them as enforced, not suggestions.
### Route categories
Every path in the desktop app falls into exactly one category. Choosing the wrong one reproduces bugs we've already fixed.
- **Session routes** — workspace-scoped pages (`/:slug/issues`, `/:slug/settings`). Rendered by the per-tab memory router under `WorkspaceRouteLayout`. These are legitimate tab destinations.
- **Transition flows** — pre-workspace / one-shot actions (create workspace, accept invite). **NOT routes.** They live as `WindowOverlay` state, dispatched when the navigation adapter sees `push('/workspaces/new')` or `push('/invite/<id>')`. The shared view (`NewWorkspacePage`, `InvitePage`) is the content; the overlay wrapper supplies platform chrome.
- **Error / stale states** — "workspace not available", tabs pointing at a revoked workspace. **NOT pages.**`WorkspaceRouteLayout` auto-heals by dropping the stale tab group from the store; the user never lands on an explicit error screen. Web keeps `NoAccessPage` (shareable URL makes the error state meaningful); desktop has no URL bar so stale = heal silently.
**Adding a new pre-workspace flow on desktop**: register a new `WindowOverlay` type in `stores/window-overlay-store.ts`. Do NOT add it to `routes.tsx`. If a shared view needs the flow on both platforms, add the route on web (`apps/web/app/(auth)/...`) AND the overlay type on desktop — the shared view component is identical.
### Workspace identity singleton
`setCurrentWorkspace(slug, uuid)` in `@multica/core/platform` is the single source of truth for "which workspace is active right now". Three consumers depend on it:
1. API client's `X-Workspace-Slug` header.
2. Zustand per-workspace storage namespace.
3. Chrome gating (`{slug && <AppSidebar />}` on desktop, similar on web).
Normally set by `WorkspaceRouteLayout` when its route mounts. Critically: **unmount does NOT clear it.** Any code that leaves workspace context (leave workspace, delete workspace, force navigation to overlay) must call `setCurrentWorkspace(null, null)` explicitly — otherwise the realtime `workspace:deleted` handler races the mutation, chrome gating stays truthy while the workspace is gone from cache, and `useWorkspaceId` throws.
### Workspace destructive operations
Leave / Delete workspace flows must follow this order:
1. Read destination from cached workspace list (no extra fetch).
2.`setCurrentWorkspace(null, null)`.
3.`navigation.push(destination)` — switch to next workspace or open new-workspace overlay.
4. THEN `await mutation.mutateAsync(workspaceId)`.
Reversing step 4 with steps 1–3 (mutate first, navigate after) causes a three-way race between the mutation's `onSettled` invalidate, the explicit `navigateAway`, and the realtime handler's `relocateAfterWorkspaceLoss` — all refetching the same `workspaces` query concurrently. One gets cancelled, bubbles as `CancelledError`, and triggers `window.location.assign` → full renderer reload / white screen.
### Tab isolation
Tabs are grouped per workspace in `stores/tab-store.ts`. The TabBar shows only the active workspace's tabs; cross-workspace tab leakage is impossible by construction (no flat global tabs array).
Cross-workspace `push(path)` is detected by the navigation adapter (`platform/navigation.tsx`) and translated into `switchWorkspace(slug, targetPath)` — NOT a navigation within the current tab's router. Don't bypass the adapter; always go through `useNavigation()` from shared code.
### Drag region (macOS window-move)
Every full-window desktop view (login, overlay, any page that covers the native title bar) needs a top drag strip so users can move the window. On macOS the traffic lights are hidden via `useImmersiveMode` in overlay-style contexts, so the drag strip also gives back that corner for pointer-drag.
**Pattern**: flex child at top, not absolute overlay.
{/* page content — interactive elements need their own "no-drag" */}
</div>
</div>
```
Why flex, not absolute: the absolute-strip + `z-index` approach relies on stacking-context hit-testing, which isn't reliable for `-webkit-app-region`. A real flex row with no siblings at that pixel is unambiguous. Height matches `MainTopBar` (48px / `h-12`) for consistency.
UX affordances (Back button, Log out button, welcome copy, invite card) belong in `packages/views/` so web and desktop render identical content. Platform chrome (drag strip, `useImmersiveMode`, tab system interaction, traffic-light accommodation) lives in desktop-only code. Violating this split always produces platform divergence — if a button exists on desktop but not on web for the same flow, it's a signal the UX escaped into platform code.
## UI/UX Rules
- Prefer shadcn components over custom implementations. Install via `pnpm ui:add <component>` from project root — adds to `packages/ui/components/ui/`. All components use Base UI primitives (`@base-ui/react`), not Radix.
The `runs` command shows all past and current executions for an issue, including running tasks. The `run-messages` command shows the detailed message log (tool calls, thinking, text, errors) for a single run. Use `--since` for efficient polling of in-progress runs.
## Projects
Projects group related issues (e.g. a sprint, an epic, a workstream). Every project
belongs to a workspace and can optionally have a lead (member or agent).
@@ -376,6 +470,63 @@ multica config set app_url https://app.example.com
multica config set workspace_id <workspace-id>
```
## Autopilot Commands
Autopilots are scheduled/triggered automations that dispatch agent tasks (either by creating an issue or by running an agent directly).
### List Autopilots
```bash
multica autopilot list
multica autopilot list --status active --output json
```
### Get Autopilot Details
```bash
multica autopilot get <id>
multica autopilot get <id> --output json # includes triggers
```
### Create / Update / Delete
```bash
multica autopilot create \
--title "Nightly bug triage"\
--description "Scan todo issues and prioritize."\
--agent "Lambda"\
--mode create_issue
multica autopilot update <id> --status paused
multica autopilot update <id> --description "New prompt"
multica autopilot delete <id>
```
`--mode` currently only accepts `create_issue` (creates a new issue on each run and assigns it to the agent). The server data model also defines `run_only`, but the daemon task path doesn't yet resolve a workspace for runs without an issue, so it's not exposed by the CLI. `--agent` accepts either a name or UUID.
### Manual Trigger
```bash
multica autopilot trigger <id> # Fires the autopilot once, returns the run
Only cron-based `schedule` triggers are currently exposed via the CLI. The data model also defines `webhook` and `api` kinds, but there is no server endpoint that fires them yet, so they're not surfaced here.
@@ -165,12 +166,12 @@ Wait 3 seconds, then verify:
multica daemon status
```
Expected output should show `running` status with detected agents (e.g. `claude`, `codex`, `opencode`, `openclaw`, `hermes`).
Expected output should show `running` status with detected agents (e.g. `claude`, `codex`, `opencode`, `openclaw`, `hermes`, `gemini`, `pi`, `cursor-agent`).
**If daemon fails to start:**
- Check logs: `multica daemon logs`
- If a port conflict occurs, the daemon may already be running under a different profile.
- If no agents are detected, ensure at least one AI CLI (`claude`, `codex`, `opencode`, `openclaw`, or`hermes`) is installed and on the `$PATH`.
- If no agents are detected, ensure at least one AI CLI (`claude`, `codex`, `opencode`, `openclaw`, `hermes`, `gemini`, `pi`, or `cursor-agent`) is installed and on the `$PATH`.
---
@@ -184,12 +185,12 @@ multica daemon status
Confirm:
1. Status is `running`
2. At least one agent is listed (e.g. `claude`, `codex`, `opencode`, `openclaw`, or`hermes`)
2. At least one agent is listed (e.g. `claude`, `codex`, `opencode`, `openclaw`, `hermes`, `gemini`, `pi`, or `cursor-agent`)
3. At least one workspace is being watched
If the agents list is empty, tell the user:
> "The Multica daemon is running but no AI agent CLIs were detected. Please install at least one supported CLI (`claude`, `codex`, `opencode`, `openclaw`, or `hermes`), then restart the daemon with `multica daemon stop && multica daemon start`."
> "The Multica daemon is running but no AI agent CLIs were detected. Please install at least one supported CLI (`claude`, `codex`, `opencode`, `openclaw`, `hermes`, `gemini`, `pi`, or `cursor-agent`), then restart the daemon with `multica daemon stop && multica daemon start`."
@@ -30,7 +30,7 @@ Turn coding agents into real teammates — assign tasks, track progress, compoun
Multica turns coding agents into real teammates. Assign issues to an agent like you'd assign to a colleague — they'll pick up the work, write code, report blockers, and update statuses autonomously.
No more copy-pasting prompts. No more babysitting runs. Your agents show up on the board, participate in conversations, and compound reusable skills over time. Think of it as open-source infrastructure for managed agents — vendor-neutral, self-hosted, and designed for human + AI teams. Works with **Claude Code**, **Codex**, **OpenClaw**, and**OpenCode**.
No more copy-pasting prompts. No more babysitting runs. Your agents show up on the board, participate in conversations, and compound reusable skills over time. Think of it as open-source infrastructure for managed agents — vendor-neutral, self-hosted, and designed for human + AI teams. Works with **Claude Code**, **Codex**, **OpenClaw**, **OpenCode**, **Hermes**, **Gemini**, **Pi**, and **Cursor Agent**.
@@ -97,7 +97,7 @@ multica setup # Connect to Multica Cloud, log in, start daemon
multica setup # Configure, authenticate, and start the daemon
```
The daemon runs in the background and auto-detects agent CLIs (`claude`, `codex`, `openclaw`, `opencode`) on your PATH.
The daemon runs in the background and auto-detects agent CLIs (`claude`, `codex`, `openclaw`, `opencode`, `hermes`, `gemini`, `pi`, `cursor-agent`) on your PATH.
### 2. Verify your runtime
@@ -107,7 +107,7 @@ Open your workspace in the Multica web app. Navigate to **Settings → Runtimes*
### 3. Create an agent
Go to **Settings → Agents** and click **New Agent**. Pick the runtime you just connected and choose a provider (Claude Code, Codex, OpenClaw, or OpenCode). Give your agent a name — this is how it will appear on the board, in comments, and in assignments.
Go to **Settings → Agents** and click **New Agent**. Pick the runtime you just connected and choose a provider (Claude Code, Codex, OpenClaw, OpenCode, Hermes, Gemini, Pi, or Cursor Agent). Give your agent a name — this is how it will appear on the board, in comments, and in assignments.
### 4. Assign your first task
@@ -158,10 +158,10 @@ See the [CLI and Daemon Guide](CLI_AND_DAEMON.md) for the full command reference
This clones the repository, starts all services via Docker Compose, installs the `multica` CLI, then configures it for localhost.
Open http://localhost:3000, log in with any email + verification code **`888888`**.
Open http://localhost:3000. To log in, configure `RESEND_API_KEY` in `.env` for email-based codes (recommended), or set `APP_ENV=development` in `.env` to enable the dev master code **`888888`**. See [Step 2 — Log In](#step-2--log-in) for details.
> **Prerequisites:** Docker and Docker Compose must be installed. The script checks for this and provides install links if missing.
>
@@ -63,9 +63,13 @@ Once ready:
### Step 2 — Log In
Open http://localhost:3000 in your browser. Enter any email address and use verification code **`888888`** to log in.
Open http://localhost:3000 in your browser. The Docker self-host stack defaults to `APP_ENV=production` (set in `docker-compose.selfhost.yml`), so the dev master code is **disabled by default** for safety on public deployments. Pick one of the following to log in:
> This master code works in all non-production environments (i.e. when `APP_ENV` is not set to `production`). For production, configure an email provider — see [Advanced Configuration](SELF_HOSTING_ADVANCED.md#email-required-for-authentication).
- **Recommended (production):** configure `RESEND_API_KEY` in `.env`, then restart the backend. Real verification codes will be sent to the email address you enter. See [Advanced Configuration → Email](SELF_HOSTING_ADVANCED.md#email-required-for-authentication).
- **Evaluation / private network:** set `APP_ENV=development` in `.env` and restart the backend. Verification code **`888888`** will then work for any email address.
- **Without configuring either:** the verification code is generated server-side and printed to the backend container logs (look for `[DEV] Verification code for ...:`). Useful for one-off testing on a single machine.
> **Warning:** do **not** set `APP_ENV=development` on a publicly reachable instance — anyone who knows an email address can then log in with `888888`.
### Step 3 — Install CLI & Start Daemon
@@ -85,6 +89,9 @@ You also need at least one AI agent CLI installed:
- [OpenClaw](https://github.com/openclaw/openclaw) (`openclaw` on PATH)
- [OpenCode](https://github.com/anomalyco/opencode) (`opencode` on PATH)
- [Hermes](https://github.com/NousResearch/hermes) (`hermes` on PATH)
- Gemini (`gemini` on PATH)
- [Pi](https://pi.dev/) (`pi` on PATH)
- [Cursor Agent](https://cursor.com/) (`cursor-agent` on PATH)
@@ -14,6 +14,15 @@ All configuration is done via environment variables. Copy `.env.example` as a st
| `JWT_SECRET` | **Must change from default.** Secret key for signing JWT tokens. Use a long random string. | `openssl rand -hex 32` |
| `FRONTEND_ORIGIN` | URL where the frontend is served (used for CORS) | `https://app.example.com` |
### Database Pool Tuning (Optional)
These have sensible defaults and only need to be set when tuning a large or constrained deployment. Precedence (highest first): env var → `pool_*` query params on `DATABASE_URL` → built-in default.
| Variable | Description | Default |
|----------|-------------|---------|
| `DATABASE_MAX_CONNS` | pgxpool max connections per pod. `pod_count × DATABASE_MAX_CONNS` should stay well below the Postgres `max_connections` ceiling. With a connection pooler (PgBouncer / RDS Proxy / Supavisor) in front, this can be raised significantly. | `25` |
| `DATABASE_MIN_CONNS` | pgxpool warm baseline connections per pod. Auto-clamped to `DATABASE_MAX_CONNS`. | `5` |
### Email (Required for Authentication)
Multica uses email-based magic link authentication via [Resend](https://resend.com).
@@ -23,7 +32,7 @@ Multica uses email-based magic link authentication via [Resend](https://resend.c
> **Note:** For local/development deployments without email configured, you can use the master verification code `888888` to log in.
> **Note:** The dev master verification code `888888` is gated by `APP_ENV != "production"`. The Docker self-host stack defaults to `APP_ENV=production` (so `888888` is disabled), which protects publicly reachable instances. For local development without email configured, set `APP_ENV=development` in your `.env` to enable `888888` — never do this on a public instance.
### Google OAuth (Optional)
@@ -44,7 +53,14 @@ For file uploads and attachments, configure S3 and CloudFront:
| `CLOUDFRONT_DOMAIN` | CloudFront distribution domain |
| `CLOUDFRONT_KEY_PAIR_ID` | CloudFront key pair ID for signed URLs |
| `COOKIE_DOMAIN` | Domain for CloudFront auth cookies |
### Cookies
| Variable | Description |
|----------|-------------|
| `COOKIE_DOMAIN` | Optional `Domain` attribute for session + CloudFront cookies. **Leave empty** for single-host deployments (localhost, LAN IP, or a single hostname). Only set it when the frontend and backend sit on different subdomains of one registered domain (e.g. `.example.com`). **Do not use an IP literal** — RFC 6265 forbids IP addresses in the cookie `Domain` attribute and browsers will drop such `Set-Cookie` headers. |
The `Secure` flag on session cookies is derived automatically from the scheme of `FRONTEND_ORIGIN`: HTTPS origins get `Secure` cookies; plain-HTTP origins (LAN / private-network self-host) get non-secure cookies so the browser can actually store them.
### Server
@@ -80,6 +96,12 @@ Agent-specific overrides:
| `MULTICA_OPENCLAW_MODEL` | Override the OpenClaw model used |
| `MULTICA_HERMES_PATH` | Custom path to the `hermes` binary |
| `MULTICA_HERMES_MODEL` | Override the Hermes model used |
| `MULTICA_GEMINI_PATH` | Custom path to the `gemini` binary |
| `MULTICA_GEMINI_MODEL` | Override the Gemini model used |
| `MULTICA_PI_PATH` | Custom path to the `pi` binary |
| `MULTICA_PI_MODEL` | Override the Pi model used |
| `MULTICA_CURSOR_PATH` | Custom path to the `cursor-agent` binary |
| `MULTICA_CURSOR_MODEL` | Override the Cursor Agent model used |
@@ -45,7 +45,7 @@ Then configure, authenticate, and start the daemon:
multica setup
```
The daemon auto-detects available agent CLIs (`claude`, `codex`, `gemini`, `openclaw`, `opencode`, `hermes`) on your PATH. When an agent is assigned a task, the daemon creates an isolated environment, runs the agent, and reports results back.
The daemon auto-detects available agent CLIs (`claude`, `codex`, `gemini`, `openclaw`, `opencode`, `hermes`, `pi`) on your PATH. When an agent is assigned a task, the daemon creates an isolated environment, runs the agent, and reports results back.
This clones the repo, starts all services, installs the CLI, and configures it for localhost. Then open http://localhost:3000 — login with any email + code **`888888`**.
This clones the repo, starts all services, installs the CLI, and configures it for localhost. Then open http://localhost:3000 and pick a login method: configure `RESEND_API_KEY` in `.env` for email-based codes (recommended), or set `APP_ENV=development` in `.env` to enable the dev master code **`888888`**. See [Step 2 — Log In](#step-2--log-in) for details.
<Callout>
If the self-host server is already running and you only need the CLI on a macOS/Linux machine, install it with Homebrew: `brew install multica-ai/tap/multica`.
@@ -64,10 +64,14 @@ If you prefer running the Docker Compose steps manually: `cp .env.example .env`,
### Step 2 — Log In
Open http://localhost:3000. Enter any email address and use verification code **`888888`** to log in.
Open http://localhost:3000. The Docker self-host stack defaults to `APP_ENV=production` (set in `docker-compose.selfhost.yml`), so the dev master code is **disabled by default** for safety on public deployments. Pick one of the following to log in:
- **Recommended (production):** configure `RESEND_API_KEY` in `.env`, then restart the backend. Real verification codes will be sent to the email address you enter. See [Configuration](#configuration) below.
- **Evaluation / private network:** set `APP_ENV=development` in `.env` and restart the backend. Verification code **`888888`** will then work for any email address.
- **Without configuring either:** the verification code is generated server-side and printed to the backend container logs (look for `[DEV] Verification code for ...:`). Useful for one-off testing on a single machine.
<Callout>
This master code works in all non-production environments (when `APP_ENV` is not set to `production`). For production, configure an email provider — see [Configuration](#configuration) below.
**Warning:** do **not** set `APP_ENV=development` on a publicly reachable instance — anyone who knows an email address can then log in with `888888`.
</Callout>
### Step 3 — Install CLI & Start Daemon
@@ -198,7 +202,14 @@ For file uploads and attachments, configure S3 and CloudFront:
| `CLOUDFRONT_DOMAIN` | CloudFront distribution domain |
| `CLOUDFRONT_KEY_PAIR_ID` | CloudFront key pair ID for signed URLs |
| `COOKIE_DOMAIN` | Domain for CloudFront auth cookies |
### Cookies
| Variable | Description |
|----------|-------------|
| `COOKIE_DOMAIN` | Optional `Domain` attribute for session + CloudFront cookies. **Leave empty** for single-host deployments (localhost, LAN IP, or a single hostname). Only set it when the frontend and backend sit on different subdomains of one registered domain (e.g. `.example.com`). **Do not use an IP literal** — RFC 6265 forbids IP addresses in the cookie `Domain` attribute and browsers will drop such `Set-Cookie` headers. |
The `Secure` flag on session cookies is derived automatically from the scheme of `FRONTEND_ORIGIN`: HTTPS origins get `Secure` cookies; plain-HTTP origins (LAN / private-network self-host) get non-secure cookies so the browser can actually store them.
@@ -11,7 +11,7 @@ Once you have the CLI installed (or signed up for [Multica Cloud](https://multic
multica setup # Configure, authenticate, and start the daemon
```
This configures the CLI, opens your browser for login, discovers your workspaces, and starts the agent daemon in the background. It auto-detects agent CLIs (`claude`, `codex`, `gemini`, `openclaw`, `opencode`, `hermes`) available on your PATH.
This configures the CLI, opens your browser for login, discovers your workspaces, and starts the agent daemon in the background. It auto-detects agent CLIs (`claude`, `codex`, `gemini`, `openclaw`, `opencode`, `hermes`, `pi`) available on your PATH.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.