The bare cache used a mirror-style fetch refspec
(+refs/heads/*:refs/heads/*) which collided with worktree-locked
refs/heads/agent/<task> branches once those branches were pushed
back to origin as PRs. git fetch aborted with "refusing to fetch
into branch ... checked out at ...", the error was swallowed as a
warning, and every subsequent checkout reused the snapshot from
the original clone.
Fix:
- Clone / migrate bare caches to a remote-tracking layout
(+refs/heads/*:refs/remotes/origin/*) so fetched heads never
land in refs/heads/*.
- Resolve the base ref from refs/remotes/origin/HEAD with a
5-level fallback (verified origin/HEAD symref to origin/main
or origin/master to the bare HEAD bridged into origin/<same>
to single-entry origin/* scan to bare HEAD for legacy caches).
- Refuse to guess when refs/remotes/origin/* has multiple
candidates and none match a known fallback, so CreateWorktree
fails loudly instead of basing work on an arbitrary branch.
- Refresh refs/remotes/origin/HEAD after every successful fetch,
not just on the legacy migration path, so a cache that was
already modern picks up an upstream default-branch change.
- Verify the primary symref target actually exists so a phantom
refs/remotes/origin/HEAD from a broken set-head does not
surface a deleted branch.
- Detect legacy caches on the fly and rewrite refspec +
refs/remotes/origin/* + refs/remotes/origin/HEAD in place so
existing clones self-heal on next use.
- Serialize per-bare-repo mutation (both Sync and CreateWorktree)
with sync.Map-backed mutexes so concurrent fetch and worktree
add on the same repo cannot race on git's own lockfiles.
- Narrow the already-exists retry to actual branch-collision
errors so a path-collision no longer silently leaks a branch
into the bare repo.
Implement the Master Agent chat feature allowing users to chat with agents
directly from a floating window, separate from the issue-based workflow.
Backend:
- New chat_session and chat_message tables (migration 033)
- Make issue_id nullable on agent_task_queue for chat tasks
- REST API: create/list/get/archive sessions, send/list messages
- EnqueueChatTask in TaskService with session_id persistence
- WS events: chat:message, chat:done
- Daemon: chat task type with separate prompt builder
- ClaimTaskByRuntime populates chat context (session, message, repos)
Frontend:
- ChatSession/ChatMessage types + API client methods
- core/chat: TanStack Query options, mutations with optimistic updates, WS updaters
- features/chat: Zustand store, ChatFab (floating button), ChatWindow with
real-time streaming via task:message events
- Mounted in dashboard layout (bottom-right corner)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(daemon): update existing worktree to latest remote on reuse
When an agent receives a new task on the same issue, the execution
environment is reused and the repo worktree already exists on disk.
Previously, `multica repo checkout` would fail because `git worktree add`
cannot create a path that already exists — so the agent worked on stale
code from the prior task.
Now `CreateWorktree` detects existing worktrees and updates them:
fetch origin, reset working tree, then checkout a new branch from the
latest remote default branch. The previous task's branch is preserved.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(daemon): propagate actual branch name and use correct ref in worktree reuse
- Return (string, error) from updateExistingWorktree so collision-retried
branch name propagates to WorktreeResult
- Use baseRef directly instead of origin/baseRef — bare clone refspec maps
remote branches to local refs, so remote-tracking refs may not exist
- Remove redundant fetch (worktree shares object store with bare clone)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1. Separate ReportTaskUsage endpoint (POST /api/daemon/tasks/{id}/usage)
so usage is captured independently of complete/fail — fixes usage loss
for failed/blocked tasks.
2. Add usage tracking for all four providers:
- Claude: already done (stream-json message.usage)
- OpenCode: extract from step_finish.part.tokens
- OpenClaw: extract from step_end.data token fields
- Codex: extract from turn/completed and task_complete usage fields
3. Remove usage from CompleteTask payload — all usage goes through the
dedicated endpoint now.
Extract token usage from Claude Code's stream-json output in real-time
during task execution, replacing the inaccurate global JSONL log scanner.
- New `task_usage` table: tracks (task_id, provider, model) level usage
- Agent SDK: parse `message.usage` from assistant messages, accumulate
per-model and return in Result
- Daemon: convert agent usage to entries, send with CompleteTask
- Server: store usage on task completion, expose workspace-level
aggregation APIs (GET /api/usage/daily, GET /api/usage/summary)
When repos are present, sub-steps c/d/e/f are now distinct instead of
having two 'c' steps. Each branch (with/without repos) now has its own
complete set of correctly lettered sub-steps.
Prepend the directory of the running multica binary to PATH in the
agent's environment variables. This fixes the issue where isolated
runtimes (e.g. Codex sandbox) cannot find the multica CLI, causing
agent tasks to fail immediately with "command not found: multica".
Closes#451
Add OpenClaw as a fourth supported agent runtime alongside Claude Code,
Codex, and OpenCode. OpenClaw CLI (`openclaw agent -p ... --output-format
stream-json`) is integrated via the same Backend interface pattern.
Changes:
- Add openclawBackend in server/pkg/agent/openclaw.go with NDJSON
event stream parsing (text, thinking, tool_call, error, step, result)
- Register "openclaw" in the agent factory (agent.go)
- Add MULTICA_OPENCLAW_PATH / MULTICA_OPENCLAW_MODEL env var detection
in daemon config
- Include "openclaw" in AGENTS.md config injection alongside codex/opencode
- Add comprehensive unit tests for all event handlers and processEvents
* refactor(cli): overhaul help output to match gh CLI style
- Add gh-style grouped help with CORE/RUNTIME/ADDITIONAL COMMANDS sections
- Use UPPERCASE section headers (USAGE, FLAGS, EXAMPLES, LEARN MORE)
- Format commands as "name: description" with automatic alignment
- Add ENVIRONMENT VARIABLES and EXAMPLES sections to root help
- Apply consistent templates to root, subcommand, and leaf commands
- Update descriptions from "Manage X" to "Work with X" for gh parity
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(execenv): add explicit instruction for agents to always use multica CLI
Agents were using curl/wget to access Multica attachment URLs directly,
which fails due to authentication. Add a prominent "Important" section
to the generated CLAUDE.md template that explicitly prohibits direct
HTTP access and instructs agents to escalate missing CLI functionality
to their workspace owner.
---------
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1. Update CLAUDE.md template to document --limit, --offset, --since
params and guide agents to use pagination when comments are large
2. Add GetJSONWithHeaders to API client; CLI now prints "Showing X of Y
comments" to stderr when paginating
3. Cap --since without --limit at 500 server-side to prevent unbounded
result sets
* fix(agent): instruct agents to use download_url for attachments
Agents were not aware of the signed vs unsigned URL distinction in
attachments, causing failures when trying to read images. Added an
Attachments section to the generated CLAUDE.md/AGENTS.md template that
tells agents to always use `download_url`. Also increased signed URL
expiry from 5 to 30 minutes to better accommodate agent processing time.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(cli): add `multica attachment download` command
Adds a dedicated CLI command for downloading attachments by ID. The
command fetches attachment metadata from the API (which returns a fresh
signed URL), downloads the file, and saves it locally. This eliminates
the need for agents to understand signed vs unsigned URLs.
Changes:
- New `multica attachment download <id>` CLI command
- New `GET /api/attachments/{id}` backend endpoint
- `DownloadFile` helper on APIClient
- Updated CLAUDE.md template to document the command
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(cli): sanitize filename and add download size limit
- Use filepath.Base on attachment filename to prevent path traversal
- Add 100MB size limit to DownloadFile (matches upload limit)
- Include response body in download error messages for debugging
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Devv <devv@Devvs-Mac-mini.local>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(daemon): add opencode as supported agent provider
Add opencode backend alongside claude and codex. The backend spawns
`opencode run --format json`, parses streaming JSON events (text,
tool_use, error, step_start/finish), and supports --prompt for system
prompts. Includes CLI detection, AGENTS.md runtime config, native skill
discovery via .config/opencode/skills/, and 21 tests covering handlers,
JSON parsing, and integration-level processEvents scenarios.
* chore: add .tool-versions to gitignore
* feat(daemon): support direct download update for non-Homebrew installs
Previously, CLI auto-update only worked for Homebrew installations. Non-brew
binaries would fail with "not installed via Homebrew". Now the daemon and
`multica update` fall back to downloading the release binary directly from
GitHub Releases when Homebrew is not detected.
Also fixes:
- Daemon restart now uses the current executable's absolute path instead of
searching PATH, ensuring the updated binary is used
- Brew installs preserve the symlink path so the new Cellar version is picked up
- Daemon startup logs now include the CLI version
- Update UI auto-clears "restarting" status after 5s to show the new version
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(cli): remove dead DetectNewBinaryPath and guard against nil latest version
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 runtime detail page was showing the agent CLI version (claude/codex)
as "CLI Version" because metadata.version stored the agent version from
agent.DetectVersion(). The multica CLI version was never sent.
Fix: daemon now sends cli_version in the registration request, server
stores it as metadata.cli_version alongside the existing agent version,
and frontend reads metadata.cli_version.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(runtime): support CLI update from web runtime page
Add the ability to update the CLI daemon from the web Runtime detail page.
When a newer version is available on GitHub Releases, an update button
appears. Clicking it sends an update command through the server to the
daemon via the heartbeat mechanism (same pattern as ping). The daemon
executes `brew upgrade`, reports the result, and restarts itself with the
new binary.
Changes across all three layers:
- Frontend: version display, GitHub latest check, UpdateSection component
- Server: UpdateStore (in-memory), heartbeat extension, 3 new endpoints
- CLI: shared update logic, daemon handleUpdate + graceful restart
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(runtime): handle 'running' status in ReportUpdateResult
The daemon sends {"status":"running"} when it starts executing the
update, but ReportUpdateResult treated any non-"completed" status as
failure — immediately marking the update as failed before brew upgrade
even ran.
Fix: use a switch statement to handle "running" as a no-op (status is
already "running" from PopPending), and also timeout running updates
after 120 seconds in case brew upgrade hangs.
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(mentions): support @mentioning issues in comments
- Extend MentionItem type to include "issue" alongside "member"/"agent"
- Add issue search (by identifier and title) to mention suggestion dropdown
- Render issue mentions with CircleDot icon in autocomplete popup
- Issue mentions serialize as [MUL-117 Title](mention://issue/id) (no @ prefix)
- Markdown renderer shows issue mentions as clickable links to /issues/:id
- Backend mentionRe regex updated to match issue mention type
* feat(mentions): auto-expand issue identifiers and add mention format to agent instructions
1. Path A — CLAUDE.md template (runtime_config.go):
Add a "## Mentions" section teaching agents the mention serialization
format for issues, members, and agents. All agents automatically
receive this via the auto-generated CLAUDE.md.
2. Approach 2 — Server-side auto-conversion (internal/mention/):
New ExpandIssueIdentifiers() utility that scans comment content for
bare issue identifiers (e.g. MUL-117) and replaces them with
[MUL-117](mention://issue/<uuid>) mention links. Skips code blocks,
inline code, and existing markdown links. Integrated into both:
- handler.CreateComment (HTTP API path)
- service.createAgentComment (agent task output path)
Users can now interrupt running agents via a "Stop" button on the live
card. The daemon polls task status every 5 seconds and kills the agent
process when cancellation is detected.
Changes:
- New CancelAgentTask SQL query and CancelTask service method
- POST /api/issues/{id}/tasks/{taskId}/cancel endpoint
- Daemon polls GetTaskStatus during execution, cancels context on match
- Frontend: Stop button on AgentLiveCard, task:cancelled WS event
renderIssueContext() now includes a "Quick Start" section with the
`multica issue get` command so agents know how to fetch issue details.
Fixes the TestPrepareDirectoryMode and TestWriteContextFiles failures.
* feat(agents): reply as thread instead of top-level comment
When an agent responds to a user comment, the reply is now nested under
the triggering comment (parent_id) instead of appearing as a separate
top-level comment. Also enables on_comment trigger by default for newly
created agents.
- Add trigger_comment_id column to agent_task_queue (migration 028)
- Pass triggering comment ID through EnqueueTaskForIssue → task → createAgentComment
- Include parent_id in WebSocket broadcast for agent comments
- Default agent creation includes both on_assign and on_comment triggers
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(cli): add --parent flag to comment add for threaded replies
The agent posts comments via the CLI, so the correct fix is giving it a
--parent flag rather than wiring trigger_comment_id through the task
infrastructure. The agent reads the comment list, decides which comment
to reply to, and passes --parent <comment-id>.
- Add --parent flag to `multica issue comment add`
- Update agent runtime instructions to explain --parent usage
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(daemon): pass trigger_comment_id to agent execution context
The agent now knows which comment triggered its task and gets an explicit
instruction to reply to it using --parent. The trigger_comment_id flows
from the DB through the claim response, daemon Task struct, and into
issue_context.md where the agent sees it.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(comments): agent replies to thread root, matching frontend behavior
When the triggering comment is itself a reply (has parent_id), resolve
to the thread root so the agent's reply stays in the same flat thread.
This matches the frontend where all replies share the top-level parent.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(cli): show parent_id and full IDs in comment list
The table output now includes a PARENT column and shows full comment IDs
(not truncated) so agents can see thread structure and use --parent.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(daemon): instruct agents to always use --output json
Agents now see explicit guidance to use --output json for all read
commands, ensuring they get structured data with full IDs and parent_id
for proper threading.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat(daemon): differentiate comment-trigger vs assign-trigger context
When triggered by a comment, the agent now gets clear instructions:
- Primary goal is to read and respond to the comment
- Do NOT change issue status just because you replied
- Only change status if explicitly requested
This prevents the agent from seeing "In Review" and stopping, since
it now understands the task is to reply, not to re-evaluate the issue.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(daemon): split workflow by trigger type in CLAUDE.md/AGENTS.md
The Workflow section in the agent's runtime config now shows a
comment-reply workflow when triggered by a comment (read comments,
find trigger, reply, don't change status) vs the full assignment
workflow (set in_progress, do work, set in_review).
Previously the agent always saw the assignment workflow, causing it
to check the issue status, see "In Review", and stop without reading
or replying to the triggering comment.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* refactor(daemon): remove duplicate workflow from issue_context.md
Workflow instructions now live only in CLAUDE.md/AGENTS.md (runtime_config.go).
issue_context.md keeps just the task data: issue ID, trigger type, and
triggering comment ID.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(task): skip duplicate comment on completion for comment-triggered tasks
When triggered by a comment, the agent posts its own reply via CLI
with --parent. The task completion path was also creating a comment
from the agent's stdout output, resulting in duplicates. Now only
assignment-triggered tasks auto-post output as a comment. Error
messages from FailTask are still posted regardless of trigger type.
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 duplicate icons in tool call rows (use chevron only for expand/collapse)
- Show detailed tool information (WebSearch queries, Agent prompts, Skill names)
- Add thinking/reasoning rows with Brain icon and expandable content
- Show tool results as separate chronological entries with previews
- Add TaskRunHistory component for viewing past agent execution logs
- Add listTasksByIssue API endpoint and task-runs route
- Support thinking content blocks in agent SDK (MessageThinking type)
- Improve callID→toolName mapping in daemon message forwarding
When an agent is working on an issue, users can now see real-time output
in the issue detail page instead of waiting for completion.
Backend:
- Add task_message table and migration for persisting agent messages
- Add POST /api/daemon/tasks/{id}/messages endpoint for daemon to report
structured messages (tool_use, tool_result, text, error) in batches
- Add GET /api/daemon/tasks/{id}/messages for catch-up after reconnect
- Add GET /api/issues/{id}/active-task to check for running tasks
- Broadcast task:message events via WebSocket
- Daemon forwards agent session messages with 500ms text throttling
Frontend:
- Add AgentLiveCard component showing live tool calls, text output,
and progress indicators with auto-scroll
- Wire into issue detail timeline with WS subscription and HTTP catch-up
- Card appears when agent is working, disappears on completion/failure
Allow running multiple daemon instances against different servers (e.g.
production and local dev) simultaneously. Each profile gets isolated
config, PID file, log file, health port, and workspaces root.
Usage:
multica login --profile dev --server-url http://localhost:8080
multica daemon start --profile dev
Default profile (no --profile flag) behavior is unchanged.
Closes MUL-42
Task execution environments were all created flat under WorkspacesRoot,
mixing tasks from different workspaces. Now tasks are nested under their
workspace ID for clearer organization and easier per-workspace cleanup.
Run syncWorkspacesFromAPI once before entering the periodic ticker
loop so newly created workspaces are discovered without the initial
30-second delay.
The daemon now periodically fetches the user's workspace list from the
API (every 30s) and adds any new workspaces to the watched config. The
existing config-watch loop then picks up the change and registers
runtimes. This fixes the issue where workspaces created after
`multica login` were not discovered until the daemon was restarted.
Enforce workspace isolation at every layer:
- Router: move RequireWorkspaceMember middleware to group level so ALL
workspace-scoped routes (issues, agents, skills, runtimes, inbox,
comments) require workspace context
- SQL: add GetXxxInWorkspace queries that filter by workspace_id,
eliminating cross-workspace data access at the query level
- Handlers: loadXForUser functions use workspace-scoped queries,
no fallback to unscoped queries
- Migration 025: add workspace_id column to comment table with backfill
- ListComments: add workspace_id filter for defense-in-depth
Fix daemon workspace mapping:
- Server returns workspace_id in task claim response (from issue)
- Daemon uses task.WorkspaceID directly instead of unreliable
workspaceIDForRuntime() local map lookup
- Remove workspaceIDForRuntime function
Fix agent/human parity:
- Comment update/delete: use resolveActor for isAuthor check so agents
can edit/delete their own comments
- Event attribution: replace hardcoded "member" with resolveActor in
agent, skill, and subscriber publish calls
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Use short task IDs (8-char prefix), task-scoped loggers for concurrent
task disambiguation, numbered tool-use at INFO level for progress
tracking, and an agent-finished summary with duration and tool count.
When agents use the multica CLI during task execution, their comments,
issue updates, and issue creations were attributed to the daemon's user
(via JWT) instead of the agent. Pass MULTICA_AGENT_ID env var from the
daemon, send X-Agent-ID header from the CLI client, and use it in
handlers to set the correct author/actor identity.
Main added execenv.Reuse() for workdir reuse across tasks on the same
issue. Our branch removed Type/BranchName/gitRoot from Environment
(repos are now checked out on demand). Resolution: keep Reuse() but
simplify it to work with the new Environment struct (no workspace type
tracking). Keep the "reused" log field from main, drop removed fields.
Agents now decide which repo to use based on issue context and check out
repos on demand via `multica repo checkout <url>`. Workspace repos are
cached locally as bare clones for fast worktree creation.
Key changes:
- Add repocache package for bare clone management (clone, fetch, worktree)
- Add `multica repo checkout` CLI command that talks to local daemon
- Add POST /repo/checkout endpoint on daemon health server
- Pass workspace repos metadata through register + task claim responses
- Remove pre-created worktrees from execenv (workdir starts empty)
- Update CLAUDE.md template to instruct agents to use `multica repo checkout`
- Pass MULTICA_DAEMON_PORT, WORKSPACE_ID, AGENT_NAME, TASK_ID env vars to agent
Previously each task created a fresh workdir via execenv.Prepare(), even
when resuming work on the same (agent, issue). This caused the agent's
session context to be out of sync with a blank code state.
Now the server returns prior_work_dir in the claim response, and the
daemon tries execenv.Reuse() first — which wraps the existing directory,
detects git worktree state, and refreshes context files. Falls back to
Prepare() if the prior workdir no longer exists. Workdirs are no longer
cleaned up after task completion so they remain available for reuse.
- Coalescing queue: use HasPendingTaskForIssue (queued/dispatched only)
instead of HasActiveTaskForIssue so comments during a running task
enqueue exactly one follow-up task that picks up all new comments.
- Stale task cleanup: runtime sweeper now fails orphaned tasks when
their runtime goes offline (daemon crash/network partition).
- Cancel-aware daemon: handleTask checks task status after execution
and discards results if the task was cancelled mid-run (e.g. reassign).
- Terminal issue guard: ClaimTaskForRuntime auto-cancels pending tasks
for done/cancelled issues instead of executing them.
- Race condition safety net: unique partial index ensures at most one
pending task per issue at the DB level.
Add an `instructions` text field to the agent model, allowing users to
define each agent's role, expertise, and working style. Instructions are
injected into CLAUDE.md as an "Agent Identity" section so the agent
knows who it is on every task execution.
- Migration 021: add instructions column to agent table
- Backend: create/update/get agent handlers support instructions
- ClaimTask response includes instructions for daemon injection
- execenv: inject instructions into CLAUDE.md meta-skill
- Frontend: add Instructions tab to agent detail panel
Store the Claude Code session ID and working directory when a task
completes. On the next task for the same (agent, issue) pair, look up
the prior session and pass --resume <session_id> to Claude Code so
the agent retains conversation context across multiple tasks on the
same issue.
Changes:
- Migration 020: add session_id and work_dir columns to agent_task_queue
- CompleteAgentTask stores session_id and work_dir on completion
- GetLastTaskSession query retrieves prior session for (agent, issue)
- ClaimTaskByRuntime handler populates prior_session_id in response
- Daemon passes ResumeSessionID through to Claude backend Execute()
- Claude backend adds --resume flag when ResumeSessionID is set
