Guru Connector (#177)

Co-authored-by: Weves <chrisweaver101@gmail.com>

backend/danswer/configs/constants.py

@@ -22,6 +22,7 @@ class DocumentSource(str, Enum):
     WEB = "web"
     GOOGLE_DRIVE = "google_drive"
     GITHUB = "github"
+    GURU = "guru"
     BOOKSTACK = "bookstack"
     CONFLUENCE = "confluence"
     SLAB = "slab"

backend/danswer/connectors/bookstack/connector.py

@@ -4,22 +4,17 @@ from collections.abc import Callable
 from datetime import datetime
 from typing import Any
 
-from bs4 import BeautifulSoup
 from danswer.configs.app_configs import INDEX_BATCH_SIZE
 from danswer.configs.constants import DocumentSource
-from danswer.configs.constants import HTML_SEPARATOR
 from danswer.connectors.bookstack.client import BookStackApiClient
 from danswer.connectors.interfaces import GenerateDocumentsOutput
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
 from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
-
+from danswer.utils.text_processing import parse_html_page_basic
-
-class BookstackClientNotSetUpError(PermissionError):
-    def __init__(self) -> None:
-        super().__init__("BookStack Client is not set up, was load_credentials called?")
 
 
 class BookstackConnector(LoadConnector, PollConnector):
@@ -135,8 +130,7 @@ class BookstackConnector(LoadConnector, PollConnector):
         page_html = (
             "<h1>" + html.escape(page_name) + "</h1>" + str(page_data.get("html"))
         )
-        soup = BeautifulSoup(page_html, "html.parser")
+        text = parse_html_page_basic(page_html)
-        text = soup.get_text(HTML_SEPARATOR)
         time.sleep(0.1)
         return Document(
             id="page:" + page_id,
@@ -148,7 +142,7 @@ class BookstackConnector(LoadConnector, PollConnector):
 
     def load_from_state(self) -> GenerateDocumentsOutput:
         if self.bookstack_client is None:
-            raise BookstackClientNotSetUpError()
+            raise ConnectorMissingCredentialError("Bookstack")
 
         return self.poll_source(None, None)
 
@@ -156,7 +150,7 @@ class BookstackConnector(LoadConnector, PollConnector):
         self, start: SecondsSinceUnixEpoch | None, end: SecondsSinceUnixEpoch | None
     ) -> GenerateDocumentsOutput:
         if self.bookstack_client is None:
-            raise BookstackClientNotSetUpError()
+            raise ConnectorMissingCredentialError("Bookstack")
 
         transform_by_endpoint: dict[
             str, Callable[[BookStackApiClient, dict], Document]

backend/danswer/connectors/confluence/connector.py

@@ -6,16 +6,16 @@ from typing import Any
 from urllib.parse import urlparse
 
 from atlassian import Confluence  # type:ignore
-from bs4 import BeautifulSoup
 from danswer.configs.app_configs import INDEX_BATCH_SIZE
 from danswer.configs.constants import DocumentSource
-from danswer.configs.constants import HTML_SEPARATOR
 from danswer.connectors.interfaces import GenerateDocumentsOutput
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
 from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
+from danswer.utils.text_processing import parse_html_page_basic
 
 # Potential Improvements
 # 1. If wiki page instead of space, do a search of all the children of the page instead of index all in the space
@@ -23,13 +23,6 @@ from danswer.connectors.models import Section
 # 3. Segment into Sections for more accurate linking, can split by headers but make sure no text/ordering is lost
 
 
-class ConfluenceClientNotSetUpError(PermissionError):
-    def __init__(self) -> None:
-        super().__init__(
-            "Confluence Client is not set up, was load_credentials called?"
-        )
-
-
 def extract_confluence_keys_from_url(wiki_url: str) -> tuple[str, str]:
     """Sample
     https://danswer.atlassian.net/wiki/spaces/1234abcd/overview
@@ -59,8 +52,7 @@ def _comment_dfs(
 ) -> str:
     for comment_page in comment_pages:
         comment_html = comment_page["body"]["storage"]["value"]
-        soup = BeautifulSoup(comment_html, "html.parser")
+        comments_str += "\nComment:\n" + parse_html_page_basic(comment_html)
-        comments_str += "\nComment:\n" + soup.get_text(HTML_SEPARATOR)
         child_comment_pages = confluence_client.get_page_child_by_type(
             comment_page["id"],
             type="comment",
@@ -101,7 +93,7 @@ class ConfluenceConnector(LoadConnector, PollConnector):
         doc_batch: list[Document] = []
 
         if self.confluence_client is None:
-            raise ConfluenceClientNotSetUpError()
+            raise ConnectorMissingCredentialError("Confluence")
 
         batch = self.confluence_client.get_all_pages_from_space(
             self.space,
@@ -116,8 +108,9 @@ class ConfluenceConnector(LoadConnector, PollConnector):
 
             if time_filter is None or time_filter(last_modified):
                 page_html = page["body"]["storage"]["value"]
-                soup = BeautifulSoup(page_html, "html.parser")
+                page_text = (
-                page_text = page.get("title", "") + "\n" + soup.get_text(HTML_SEPARATOR)
+                    page.get("title", "") + "\n" + parse_html_page_basic(page_html)
+                )
                 comment_pages = self.confluence_client.get_page_child_by_type(
                     page["id"],
                     type="comment",
@@ -146,7 +139,7 @@ class ConfluenceConnector(LoadConnector, PollConnector):
 
     def load_from_state(self) -> GenerateDocumentsOutput:
         if self.confluence_client is None:
-            raise ConfluenceClientNotSetUpError()
+            raise ConnectorMissingCredentialError("Confluence")
 
         start_ind = 0
         while True:
@@ -162,7 +155,7 @@ class ConfluenceConnector(LoadConnector, PollConnector):
         self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch
     ) -> GenerateDocumentsOutput:
         if self.confluence_client is None:
-            raise ConfluenceClientNotSetUpError()
+            raise ConnectorMissingCredentialError("Confluence")
 
         start_time = datetime.fromtimestamp(start, tz=timezone.utc)
         end_time = datetime.fromtimestamp(end, tz=timezone.utc)

backend/danswer/connectors/danswer_jira/connector.py

@@ -9,6 +9,7 @@ from danswer.connectors.interfaces import GenerateDocumentsOutput
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
 from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
 from danswer.utils.logger import setup_logger
@@ -97,9 +98,7 @@ class JiraConnector(LoadConnector, PollConnector):
 
     def load_from_state(self) -> GenerateDocumentsOutput:
         if self.jira_client is None:
-            raise PermissionError(
+            raise ConnectorMissingCredentialError("Jira")
-                "Jira Client is not set up, was load_credentials called?"
-            )
 
         start_ind = 0
         while True:
@@ -121,9 +120,7 @@ class JiraConnector(LoadConnector, PollConnector):
         self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch
     ) -> GenerateDocumentsOutput:
         if self.jira_client is None:
-            raise PermissionError(
+            raise ConnectorMissingCredentialError("Jira")
-                "Jira Client is not set up, was load_credentials called?"
-            )
 
         start_date_str = datetime.fromtimestamp(start, tz=timezone.utc).strftime(
             "%Y-%m-%d %H:%M"

backend/danswer/connectors/factory.py

@@ -9,6 +9,7 @@ from danswer.connectors.file.connector import LocalFileConnector
 from danswer.connectors.github.connector import GithubConnector
 from danswer.connectors.google_drive.connector import GoogleDriveConnector
 from danswer.connectors.notion.connector import NotionConnector
+from danswer.connectors.guru.connector import GuruConnector
 from danswer.connectors.interfaces import BaseConnector
 from danswer.connectors.interfaces import EventConnector
 from danswer.connectors.interfaces import LoadConnector
@@ -46,6 +47,7 @@ def identify_connector_class(
         DocumentSource.PRODUCTBOARD: ProductboardConnector,
         DocumentSource.SLAB: SlabConnector,
         DocumentSource.NOTION: NotionConnector,
+        DocumentSource.GURU: GuruConnector,
     }
     connector_by_source = connector_map.get(source, {})
 

backend/danswer/connectors/github/connector.py

@@ -6,6 +6,7 @@ from danswer.configs.app_configs import INDEX_BATCH_SIZE
 from danswer.configs.constants import DocumentSource
 from danswer.connectors.interfaces import GenerateDocumentsOutput
 from danswer.connectors.interfaces import LoadConnector
+from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
 from danswer.utils.logger import setup_logger
@@ -48,9 +49,7 @@ class GithubConnector(LoadConnector):
 
     def load_from_state(self) -> GenerateDocumentsOutput:
         if self.github_client is None:
-            raise PermissionError(
+            raise ConnectorMissingCredentialError("GitHub")
-                "Github Client is not set up, was load_credentials called?"
-            )
         repo = self.github_client.get_repo(f"{self.repo_owner}/{self.repo_name}")
         pull_requests = repo.get_pulls(state=self.state_filter)
         for pr_batch in get_pr_batches(pull_requests, self.batch_size):

backend/danswer/connectors/guru/connector.py

@@ -0,0 +1,110 @@
+import json
+from datetime import datetime
+from datetime import timezone
+from typing import Any
+
+import requests
+from danswer.configs.app_configs import INDEX_BATCH_SIZE
+from danswer.configs.constants import DocumentSource
+from danswer.connectors.interfaces import GenerateDocumentsOutput
+from danswer.connectors.interfaces import LoadConnector
+from danswer.connectors.interfaces import PollConnector
+from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
+from danswer.connectors.models import Document
+from danswer.connectors.models import Section
+from danswer.utils.logger import setup_logger
+from danswer.utils.text_processing import parse_html_page_basic
+
+# Potential Improvements
+# 1. Support fetching per collection via collection token (configured at connector creation)
+
+GURU_API_BASE = "https://api.getguru.com/api/v1/"
+GURU_QUERY_ENDPOINT = GURU_API_BASE + "search/query"
+GURU_CARDS_URL = "https://app.getguru.com/card/"
+logger = setup_logger()
+
+
+def unixtime_to_guru_time_str(unix_time: SecondsSinceUnixEpoch) -> str:
+    date_obj = datetime.fromtimestamp(unix_time, tz=timezone.utc)
+    date_str = date_obj.strftime("%Y-%m-%dT%H:%M:%S.%f")[:-3]
+    tz_str = date_obj.strftime("%z")
+    return date_str + tz_str
+
+
+class GuruConnector(LoadConnector, PollConnector):
+    def __init__(
+        self,
+        batch_size: int = INDEX_BATCH_SIZE,
+        guru_user: str | None = None,
+        guru_user_token: str | None = None,
+    ) -> None:
+        self.batch_size = batch_size
+        self.guru_user = guru_user
+        self.guru_user_token = guru_user_token
+
+    def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
+        self.guru_user = credentials["guru_user"]
+        self.guru_user_token = credentials["guru_user_token"]
+        return None
+
+    def _process_cards(
+        self, start_str: str | None = None, end_str: str | None = None
+    ) -> GenerateDocumentsOutput:
+        if self.guru_user is None or self.guru_user_token is None:
+            raise ConnectorMissingCredentialError("Guru")
+
+        doc_batch: list[Document] = []
+
+        session = requests.Session()
+        session.auth = (self.guru_user, self.guru_user_token)
+
+        params: dict[str, str | int] = {"maxResults": self.batch_size}
+
+        if start_str is not None and end_str is not None:
+            params["q"] = f"lastModified >= {start_str} AND lastModified < {end_str}"
+
+        current_url = GURU_QUERY_ENDPOINT  # This is how they handle pagination, a different url will be provided
+        while True:
+            response = session.get(current_url, params=params)
+
+            if response.status_code == 204:
+                break
+
+            cards = json.loads(response.text)
+            for card in cards:
+                title = card["preferredPhrase"]
+                link = GURU_CARDS_URL + card["slug"]
+                content_text = title + "\n" + parse_html_page_basic(card["content"])
+
+                doc_batch.append(
+                    Document(
+                        id=card["id"],
+                        sections=[Section(link=link, text=content_text)],
+                        source=DocumentSource.GURU,
+                        semantic_identifier=title,
+                        metadata={},
+                    )
+                )
+
+                if len(doc_batch) >= self.batch_size:
+                    yield doc_batch
+                    doc_batch = []
+
+            if not hasattr(response, "links") or not response.links:
+                break
+            current_url = response.links["next-page"]["url"]
+
+        if doc_batch:
+            yield doc_batch
+
+    def load_from_state(self) -> GenerateDocumentsOutput:
+        return self._process_cards()
+
+    def poll_source(
+        self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch
+    ) -> GenerateDocumentsOutput:
+        start_time = unixtime_to_guru_time_str(start)
+        end_time = unixtime_to_guru_time_str(end)
+
+        return self._process_cards(start_time, end_time)

backend/danswer/connectors/models.py

@@ -6,6 +6,14 @@ from danswer.configs.constants import DocumentSource
 from pydantic import BaseModel
 
 
+class ConnectorMissingCredentialError(PermissionError):
+    def __init__(self, connector_name: str) -> None:
+        connector_name = connector_name or "Unknown"
+        super().__init__(
+            f"{connector_name} connector missing credentials, was load_credentials called?"
+        )
+
+
 @dataclass
 class Section:
     link: str

backend/danswer/connectors/slab/connector.py

@@ -13,6 +13,7 @@ from danswer.connectors.interfaces import GenerateDocumentsOutput
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
 from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
 from danswer.utils.logger import setup_logger
@@ -24,11 +25,6 @@ SLAB_API_URL = "https://api.slab.com/v1/graphql"
 logger = setup_logger()
 
 
-class SlabBotTokenNotFoundError(PermissionError):
-    def __init__(self) -> None:
-        super().__init__("Slab Bot Token not found, was load_credentials called?")
-
-
 def run_graphql_request(
     graphql_query: dict, bot_token: str, max_tries: int = SLAB_GRAPHQL_MAX_TRIES
 ) -> str:
@@ -179,7 +175,7 @@ class SlabConnector(LoadConnector, PollConnector):
         doc_batch: list[Document] = []
 
         if self.slab_bot_token is None:
-            raise SlabBotTokenNotFoundError()
+            raise ConnectorMissingCredentialError("Slab")
 
         all_post_ids: list[str] = get_all_post_ids(self.slab_bot_token)
 

backend/danswer/connectors/slack/connector.py

@@ -12,6 +12,7 @@ from danswer.connectors.interfaces import GenerateDocumentsOutput
 from danswer.connectors.interfaces import LoadConnector
 from danswer.connectors.interfaces import PollConnector
 from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
 from danswer.connectors.models import Document
 from danswer.connectors.models import Section
 from danswer.connectors.slack.utils import get_message_link
@@ -285,9 +286,7 @@ class SlackPollConnector(PollConnector):
         self, start: SecondsSinceUnixEpoch, end: SecondsSinceUnixEpoch
     ) -> GenerateDocumentsOutput:
         if self.client is None:
-            raise PermissionError(
+            raise ConnectorMissingCredentialError("Slack")
-                "Slack Client is not set up, was load_credentials called?"
-            )
 
         documents: list[Document] = []
         for document in get_all_docs(

backend/danswer/utils/text_processing.py

@@ -1,3 +1,7 @@
+from bs4 import BeautifulSoup
+from danswer.configs.constants import HTML_SEPARATOR
+
+
 def clean_model_quote(quote: str, trim_length: int) -> str:
     quote_clean = quote.strip()
     if quote_clean[0] == '"':
@@ -29,3 +33,8 @@ def shared_precompare_cleanup(text: str) -> str:
     text = text.replace("-", "")
 
     return text
+
+
+def parse_html_page_basic(text: str) -> str:
+    soup = BeautifulSoup(text, "html.parser")
+    return soup.get_text(HTML_SEPARATOR)

web/public/Guru.svg

@@ -0,0 +1 @@
+<?xml version="1.0" encoding="UTF-8"?><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 162.31 162.01"><defs><style>.d{fill:#fff;}.e{fill:#080b0e;}</style></defs><g id="a"/><g id="b"><g id="c"><g><path class="e" d="M162.28,96.29c.57,20.21-8.19,35.32-25.42,45.02-23.91,13.45-49.54,21.53-77.3,20.64-15.64-.5-28.15-8.01-36.35-21.09C7.51,115.82-.94,88.36,.08,58.64c.54-15.6,8.6-27.88,21.78-35.94C46.67,7.52,73.69-.9,102.98,.08c15.87,.53,28.12,8.76,36.46,22.05,14.21,22.64,20.84,47.73,22.84,74.16Z"/><g><path class="d" d="M92.34,111.57c-6.26,2.28-12.96,2.67-18.88,1.08-14.71-3.95-25.39-17.48-25.39-32.16,0-11.7,5.26-19.36,9.68-23.73,6.08-6.01,14.53-9.58,22.67-9.58,.1,0,.21,0,.31,0,15.52,.17,25.35,11.89,25.44,12,1.35,1.66,3.78,1.9,5.43,.56,1.65-1.34,1.91-3.78,.56-5.43-.49-.6-12.13-14.64-31.35-14.85-10.38-.06-20.79,4.19-28.49,11.8-7.72,7.63-11.97,18.01-11.97,29.22,0,18.11,13.08,34.78,31.11,39.62,3.05,.82,6.25,1.22,9.5,1.22,4.68,0,9.47-.84,14.04-2.51,11.87-4.34,23.47-15.76,23.05-28.96-2.62,1.92-5.47,3.67-8.47,5.33-2.31,7.25-9.61,13.6-17.23,16.38Z"/><path class="d" d="M121.5,71.35c-.07,.07-3.39,3.63-8.36,7.2-5.16-5.03-13.38-7.42-19.32-7.42-4.87,0-8.59,1.28-11.07,3.79-1.95,1.98-2.95,4.57-2.91,7.49,.08,5.25,3.85,10.93,14.16,11.16,5.75,.14,11.38-1.72,16.29-4.24,3.43-1.74,5.9-3.48,7.01-4.27,5.76-4.07,9.56-8.13,9.88-8.48,1.45-1.57,1.35-4.01-.22-5.46-1.56-1.44-4-1.35-5.46,.21Zm-27.34,14.5c-2.99-.07-6.56-.74-6.6-3.56-.02-1.23,.52-1.78,.69-1.95,.55-.56,2-1.49,5.57-1.49,3.78,0,8.83,1.45,12.33,3.95-3.76,1.85-7.81,3.14-11.99,3.05Z"/></g></g></g></g></svg>

web/src/app/admin/connectors/guru/page.tsx

@@ -0,0 +1,233 @@
+"use client";
+
+import * as Yup from "yup";
+import { GuruIcon, TrashIcon } from "@/components/icons/icons";
+import { TextFormField } from "@/components/admin/connectors/Field";
+import { HealthCheckBanner } from "@/components/health/healthcheck";
+import { CredentialForm } from "@/components/admin/connectors/CredentialForm";
+import {
+  Credential,
+  ProductboardConfig,
+  ConnectorIndexingStatus,
+  GuruConfig,
+  GuruCredentialJson,
+} from "@/lib/types";
+import useSWR, { useSWRConfig } from "swr";
+import { fetcher } from "@/lib/fetcher";
+import { LoadingAnimation } from "@/components/Loading";
+import { deleteCredential, linkCredential } from "@/lib/credential";
+import { ConnectorForm } from "@/components/admin/connectors/ConnectorForm";
+import { ConnectorsTable } from "@/components/admin/connectors/table/ConnectorsTable";
+import { usePopup } from "@/components/admin/connectors/Popup";
+
+const Main = () => {
+  const { popup, setPopup } = usePopup();
+
+  const { mutate } = useSWRConfig();
+  const {
+    data: connectorIndexingStatuses,
+    isLoading: isConnectorIndexingStatusesLoading,
+    error: isConnectorIndexingStatusesError,
+  } = useSWR<ConnectorIndexingStatus<any>[]>(
+    "/api/manage/admin/connector/indexing-status",
+    fetcher
+  );
+  const {
+    data: credentialsData,
+    isLoading: isCredentialsLoading,
+    isValidating: isCredentialsValidating,
+    error: isCredentialsError,
+  } = useSWR<Credential<any>[]>("/api/manage/credential", fetcher);
+
+  if (
+    isConnectorIndexingStatusesLoading ||
+    isCredentialsLoading ||
+    isCredentialsValidating
+  ) {
+    return <LoadingAnimation text="Loading" />;
+  }
+
+  if (isConnectorIndexingStatusesError || !connectorIndexingStatuses) {
+    return <div>Failed to load connectors</div>;
+  }
+
+  if (isCredentialsError || !credentialsData) {
+    return <div>Failed to load credentials</div>;
+  }
+
+  const guruConnectorIndexingStatuses: ConnectorIndexingStatus<GuruConfig>[] =
+    connectorIndexingStatuses.filter(
+      (connectorIndexingStatus) =>
+        connectorIndexingStatus.connector.source === "guru"
+    );
+  const guruCredential: Credential<GuruCredentialJson> = credentialsData.filter(
+    (credential) => credential.credential_json?.guru_user
+  )[0];
+
+  return (
+    <>
+      {popup}
+      <p className="text-sm">
+        This connector allows you to sync all your Guru Cards into Danswer.
+      </p>
+
+      <h2 className="font-bold mb-2 mt-6 ml-auto mr-auto">
+        Step 1: Provide your Credentials
+      </h2>
+
+      {guruCredential ? (
+        <>
+          <div className="flex mb-1 text-sm">
+            <p className="my-auto">Existing Access Token: </p>
+            <p className="ml-1 italic my-auto max-w-md truncate">
+              {guruCredential.credential_json?.guru_user_token}
+            </p>
+            <button
+              className="ml-1 hover:bg-gray-700 rounded-full p-1"
+              onClick={async () => {
+                if (guruConnectorIndexingStatuses.length > 0) {
+                  setPopup({
+                    type: "error",
+                    message:
+                      "Must delete all connectors before deleting credentials",
+                  });
+                  return;
+                }
+                await deleteCredential(guruCredential.id);
+                mutate("/api/manage/credential");
+              }}
+            >
+              <TrashIcon />
+            </button>
+          </div>
+        </>
+      ) : (
+        <>
+          <p className="text-sm">
+            To use the Guru connector, first follow the guide{" "}
+            <a
+              className="text-blue-500"
+              href="https://help.getguru.com/s/article/how-to-obtain-your-api-credentials"
+              target="_blank"
+            >
+              here
+            </a>{" "}
+            to generate a User Token.
+          </p>
+          <div className="border-solid border-gray-600 border rounded-md p-6 mt-2">
+            <CredentialForm<GuruCredentialJson>
+              formBody={
+                <>
+                  <TextFormField name="guru_user" label="Username:" />
+                  <TextFormField
+                    name="guru_user_token"
+                    label="User Token:"
+                    type="password"
+                  />
+                </>
+              }
+              validationSchema={Yup.object().shape({
+                guru_user: Yup.string().required(
+                  "Please enter your Guru username"
+                ),
+                guru_user_token: Yup.string().required(
+                  "Please enter your Guru access token"
+                ),
+              })}
+              initialValues={{
+                guru_user: "",
+                guru_user_token: "",
+              }}
+              onSubmit={(isSuccess) => {
+                if (isSuccess) {
+                  mutate("/api/manage/credential");
+                }
+              }}
+            />
+          </div>
+        </>
+      )}
+
+      <h2 className="font-bold mb-2 mt-6 ml-auto mr-auto">
+        Step 2: Start indexing!
+      </h2>
+      {guruCredential ? (
+        !guruConnectorIndexingStatuses.length ? (
+          <>
+            <p className="text-sm mb-2">
+              Click the button below to start indexing! We will pull the latest
+              features, components, and products from Guru every <b>10</b>{" "}
+              minutes.
+            </p>
+            <div className="flex">
+              <ConnectorForm<GuruConfig>
+                nameBuilder={() => "GuruConnector"}
+                source="guru"
+                inputType="poll"
+                formBody={null}
+                validationSchema={Yup.object().shape({})}
+                initialValues={{}}
+                refreshFreq={10 * 60} // 10 minutes
+                onSubmit={async (isSuccess, responseJson) => {
+                  if (isSuccess && responseJson) {
+                    await linkCredential(responseJson.id, guruCredential.id);
+                    mutate("/api/manage/admin/connector/indexing-status");
+                  }
+                }}
+              />
+            </div>
+          </>
+        ) : (
+          <>
+            <p className="text-sm mb-2">
+              Guru connector is setup! We are pulling the latest cards from Guru
+              every <b>10</b> minutes.
+            </p>
+            <ConnectorsTable<GuruConfig, GuruCredentialJson>
+              connectorIndexingStatuses={guruConnectorIndexingStatuses}
+              liveCredential={guruCredential}
+              getCredential={(credential) => {
+                return (
+                  <div>
+                    <p>{credential.credential_json.guru_user}</p>
+                  </div>
+                );
+              }}
+              onCredentialLink={async (connectorId) => {
+                if (guruCredential) {
+                  await linkCredential(connectorId, guruCredential.id);
+                  mutate("/api/manage/admin/connector/indexing-status");
+                }
+              }}
+              onUpdate={() =>
+                mutate("/api/manage/admin/connector/indexing-status")
+              }
+            />
+          </>
+        )
+      ) : (
+        <>
+          <p className="text-sm">
+            Please provide your access token in Step 1 first! Once done with
+            that, you can then start indexing all your Guru cards.
+          </p>
+        </>
+      )}
+    </>
+  );
+};
+
+export default function Page() {
+  return (
+    <div className="mx-auto container">
+      <div className="mb-4">
+        <HealthCheckBanner />
+      </div>
+      <div className="border-solid border-gray-600 border-b mb-4 pb-2 flex">
+        <GuruIcon size="32" />
+        <h1 className="text-3xl font-bold pl-2">Guru</h1>
+      </div>
+      <Main />
+    </div>
+  );
+}

web/src/app/admin/layout.tsx

@@ -9,6 +9,7 @@ import {
   KeyIcon,
   BookstackIcon,
   ConfluenceIcon,
+  GuruIcon,
   FileIcon,
   JiraIcon,
   SlabIcon,
@@ -131,6 +132,15 @@ export default async function AdminLayout({
                   ),
                   link: "/admin/connectors/notion",
                 },
+                {
+                  name: (
+                    <div className="flex">
+                      <GuruIcon size={16} />
+                      <div className="ml-1">Guru</div>
+                    </div>
+                  ),
+                  link: "/admin/connectors/guru",
+                },
                 {
                   name: (
                     <div className="flex">

web/src/components/icons/icons.tsx

@@ -18,6 +18,7 @@ import { FaFile, FaGlobe } from "react-icons/fa";
 import Image from "next/image";
 import jiraSVG from "../../../public/Jira.svg";
 import confluenceSVG from "../../../public/Confluence.svg";
+import guruIcon from "../../../public/Guru.svg";
 
 interface IconProps {
   size?: number;
@@ -237,3 +238,15 @@ export const NotionIcon = ({
     </div>
   );
 };
+
+export const GuruIcon = ({
+  size = 16,
+  className = defaultTailwindCSS,
+}: IconProps) => (
+  <div
+    style={{ width: `${size}px`, height: `${size}px` }}
+    className={`w-[${size}px] h-[${size}px] ` + className}
+  >
+    <Image src={guruIcon} alt="Logo" width="96" height="96" />
+  </div>
+);

web/src/components/search/Filters.tsx

@@ -14,6 +14,7 @@ const sources: Source[] = [
   { displayName: "Slab", internalName: "slab" },
   { displayName: "Github PRs", internalName: "github" },
   { displayName: "Web", internalName: "web" },
+  { displayName: "Guru", internalName: "guru" },
   { displayName: "File", internalName: "file" },
   { displayName: "Notion", internalName: "notion" },
 ];

web/src/components/source.tsx

@@ -6,6 +6,7 @@ import {
   GithubIcon,
   GlobeIcon,
   GoogleDriveIcon,
+  GuruIcon,
   JiraIcon,
   NotionIcon,
   ProductboardIcon,
@@ -87,6 +88,12 @@ export const getSourceMetadata = (sourceType: ValidSources): SourceMetadata => {
         displayName: "Notion",
         adminPageLink: "/admin/connectors/notion",
       };
+    case "guru":
+      return {
+        icon: GuruIcon,
+        displayName: "Guru",
+        adminPageLink: "/admin/connectors/guru",
+      };
     default:
       throw new Error("Invalid source type");
   }

web/src/lib/types.ts

@@ -17,8 +17,9 @@ export type ValidSources =
   | "jira"
   | "productboard"
   | "slab"
-  | "file"
+  | "notion"
-  | "notion";
+  | "guru"
+  | "file";
 export type ValidInputTypes = "load_state" | "poll" | "event";
 
 // CONNECTORS
@@ -72,6 +73,8 @@ export interface SlabConfig {
   base_url: string;
 }
 
+export interface GuruConfig {}
+
 export interface FileConfig {
   file_locations: string[];
 }
@@ -139,3 +142,7 @@ export interface SlabCredentialJson {
 export interface NotionCredentialJson {
   notion_integration_token: string;
 }
+export interface GuruCredentialJson {
+  guru_user: string;
+  guru_user_token: string;
+}