- Editing a connection-backed tool validates the config before writing
new credentials, so a rejected edit no longer rotates the connection's
key and resumes its sources.
- An MCP edit keeps its previous connection only when the caller owns it
and it signs in the way the tool now does.
Covers creating and editing tools whose key lives on a connection, the
MCP action refresh, and the auth, ownership and error branches of the
connections API. Editing a tool's key now also recovers a connection
whose stored credentials can no longer be decrypted.
Editing an MCP server's URL kept the previous connection when no new
one could be stored (default encryption key), so its key would reach the
new server. The save route now keeps a previous connection only for the
same server, and the executor refuses an MCP tool whose connection was
stored for another server.
- Connection groups sit between the default and custom tools; a second
sort that did not know connection labels moved them to the top.
- Upload picks the first saved account when connections load after the
modal opened on S3 or Reddit, without overriding the user's choice.
- An MCP connection's tokens only go to its own server, and a
connection_id can no longer come from a client: the MCP test/save
routes drop it and the tool executor uses only the resolved one.
- API-key connections are reused only when they hold the same
credentials. Two keys that share a hint or a label get separate
connections ("…abcd (2)"), at runtime and in the 0038 backfill.
- Reconnecting a connection whose secrets cannot be decrypted no longer
flags it from a second transaction while holding its row lock.
- Connection setup validates the sync request before claiming its
Idempotency-Key, so a corrected retry is queued.
- A request that names no connection resolves to none without opening
a database connection; fixes two tests that reached a real database.
- Guides > Connectors: using connections, sharing modes, attribution,
and admin setup (encryption key and rotation, redirect URIs, Google
Drive, SharePoint, Confluence, S3, MCP presets).
- Integration pages register CONNECTOR_REDIRECT_BASE_URI as-is; the
?provider= suffix never matched what the backend sends.
VITE_GOOGLE_CLIENT_ID is optional.
- Upgrading: set ENCRYPTION_SECRET_KEY before migrating multi-user installs.
Scheduled syncs checked the connection's status but not the admin
policy, so a disabled connector kept pulling data. Connector sources now
return "disabled" and S3/Reddit syncs get no credentials until the
connector is turned back on.
- Tool chips and the streaming status name the service: "Searched
Notion", "Used Linear: create issue". Citations from a connector
source read "From Google Drive".
- The share dialog for a connection-backed tool asks whose account team
members use. Owner mode warns which account they act as and needs an
explicit OK when the tool can write; an admin-forced mode is locked.
- Admin > Connectors: enable or disable each connector, force a sharing
mode, allow custom MCP servers, and a setup guide per OAuth connector
with its redirect URI and the server settings still missing. Warns
when credentials use the default encryption key.
- The composer and the agent builder group tools by connection, and the
Sources and Tools popovers link to Connect more.
- MCP presets open the server form titled after the service.
Downgrade removes API-key connections and assumed every tool still had
its v1 copy. Tools added through the wizard after the upgrade never had
one, and a reconnect can change the key after the backfill, so write the
connection's current credentials back to each linked tool first.
Six vendor-run MCP servers (Notion, Linear, Atlassian, Sentry, Asana,
Stripe) join the catalog as presets from docsgpt/connectors/presets/
mcp.yaml; existing connections to those servers show under them.
connector_policies lets an admin turn a connector off or force whose
account every share uses, and app_metadata's
connectors.allow_custom_mcp turns custom MCP servers off. Both are
enforced on the server: new connections, OAuth sign-ins, MCP test and
save, and the tools of a disabled connector stop resolving.
GET/PUT /api/admin/connectors reads and changes them.
A shared tool's owner picks owner or member credentials with
PUT /api/connections/tools/<id>/credential-mode, and a member running
the owner's account always confirms write actions.
Tool-call events and retrieved chunks name the connector they came
from (key and display name, never the account), and tool calls keep
those fields when the conversation is reloaded.
Every entry point (the Connectors page and drawer, Add Source, Add
Tool, the chat) opens one connect wizard: sign in or paste the
credentials, choose what to sync (content services only, skippable),
then a summary with Try it in chat. Tool services create their tools on
sign-in with write actions needing approval, and each action's
permission (Always allow, Needs approval, Off) can be changed in the
wizard or the drawer. The drawer also reconnects, syncs more content,
refreshes an MCP server's tools and removes a connection, keeping or
deleting what it feeds.
The file pickers work from a connection id and fetch a short-lived
picker token when they need one; no session token is kept in
localStorage any more, and any left by an older build is claimed once
and removed. S3 and Reddit reuse the keys stored on a connection.
A tool whose account needs signing in pauses on a Connect card in the
chat; after connecting, Continue resumes the call. A connection whose
sign-in stops working raises a Reconnect toast. The Connectors page is
hidden when the backend does not report connectors_enabled.
Migration 0038 moves every stored secret (OAuth tokens, MCP OAuth
tokens and client registrations, API keys) into the connection's
encrypted envelope, links API-key tools to one connection per distinct
credential, allows several accounts per provider, and adds
credential_mode to sources and tools. OAuth MCP tools keep resolving
each member's own token, as they did before.
docsgpt.connectors.service is now the only reader of OAuth tokens:
get_valid_token_info refreshes under a row lock and persists rotated
refresh tokens, and a revoked grant flags the connection, pauses its
sources and notifies the owner. Loaders build from a connection
(BaseConnectorLoader.from_connection), so scheduled sync covers Drive,
SharePoint and Confluence sources with no browser. S3 and Reddit keys
stay on the connection instead of in remote_data.
New endpoints: POST /api/connections, /setup, /reconnect,
/picker-token, /claim, DELETE /api/connections/<id>, per-action
permissions and MCP refresh-tools. Upload, file listing, sync and
validate-session take a connection_id; session tokens keep working for
this release. The tool executor reads credentials from the resolved
connection (owner or member mode) and pauses on a Connect card when a
connection needs signing in. docsgpt connectors reencrypt rewrites
stored credentials after a key rotation.
Adds a v2 credential envelope next to the v1 tool-secret helpers:
AES-256-GCM, a master key derived once per process from
ENCRYPTION_SECRET_KEY, and a per-record key from HKDF over the owner's
id, which is also the associated data, so a blob moved onto another
user's row does not decrypt. The envelope names its key, so
ENCRYPTION_SECRET_KEY_PREVIOUS keeps old rows readable during a
rotation.
Log redaction now also covers token_info, tokens and client_info, and
the API warns at startup when the public default key is in use.
Add Source splits into Upload and web, and From a connection. Connector
tiles are always shown and say whether an account is connected or the
server still needs setup, instead of disappearing when a VITE_ client
id is unset; without the Google Picker key, Drive falls back to the
server-side file browser. The name follows the first picked item and
the button reads Add source.
Add Tool splits into Built in and From a service. The MCP and API tool
cards move to Connectors > Add custom connector.
Settings > Workspace gets a Connectors page between Sources and Tools:
a searchable catalog filtered by category, one card per service with
its capabilities (Sync, Read, Write) and state (Connect, Connected,
Reconnect, Needs admin setup, Disabled by admin, Custom). A card opens
a drawer listing the accounts, the sources each syncs and the tools
each provides, with Reconnect and Disconnect.
Tools and Sources show which connection an item comes from and when
its sync is paused. Brand icons now draw in currentColor so they
follow the theme, and all new copy is translated in every locale.
connector_sessions already holds one row per signed-in account or MCP
server. Migration 0038 names each row (connector_key, auth_kind,
display_name, account_label) and lets sources and tools point at the
row they use through connection_id, backfilling the links it can prove.
docsgpt/connectors/catalog.py declares every service DocsGPT connects
to, with the server settings it needs and what a connection provides.
New endpoints: GET /api/connectors/catalog, GET /api/connections,
GET /api/connections/<id> and POST /api/connections/<id>/disconnect.
/api/available_tools now groups tools into built_in, service and
custom, and source listings carry connectionId.
chunks is a total per request, split across the attached sources, so an
agent with two sources and the default of 2 got a single chunk from each,
and its answers changed with whichever chunk won. 6 gives three per
source for about 4-5k more input tokens per retrieval turn.
Every literal default moves from 2 to 6: the request default, the
retrievers, the internal search tool, workflow agent nodes, scheduled and
headless runs, agent create/update/import, the source retrieval config and
the frontend forms. Existing agents and sources keep what they store; a
source saved with chunks=2 now counts as configured at 2, which is pinned
by a test.
Headless runs also read chunks=0 as unset (`or 2`), so an agent with
retrieval switched off retrieved anyway on scheduled runs; 0 now stays 0.
A display formula wider than the column scrolls in its own box, as code
and tables do, and a formula in a table header is not uppercased: n and N
are different variables.
Inline \( \) math rendered as a literal $x$ (the rewrite produced single
dollars, which remark-math is told to ignore), a one-line \[ \] shrank to
inline math, model-written $x$ never rendered, and the citation pass
linked [1] inside formulas.
- \( \) and \[ \] are micromark constructs: inline anywhere, display when
\[ starts a line and its \] ends one. A line such as "\[a\] and more"
stays inline instead of swallowing the rest of the answer, which is why
this does not alias micromark-extension-llm-math.
- $x$ follows Pandoc's rules (ported from LibreChat, MIT), so prices stay
text.
- A paragraph holding only $$ or \[ math, and anything KaTeX can only set in
display mode (\tag, align, ...), becomes display math.
- mhchem is loaded for \ce and \pu.
- Citations are linked on the syntax tree, so code, math, link text and
escaped brackets are never touched.
- While streaming, open emphasis, code and $$ / \[ math are closed (remend
plus a code-aware math pass) and a formula KaTeX cannot set yet shows in
the muted colour. Each top-level block renders on its own and is
memoised, so only the growing block is parsed and typeset again.
The remove button's label is "Remove attachment" and doesn't say which
file. Its failure description now reads "<file>: <reason>", the same text
as the tooltip.
Follow-ups to moving the failure reason into a tooltip:
- The tooltip now reads "<file>: <reason>". The name is truncated on the
chip, and two long names can look the same.
- A polite role="status" region, always mounted and screen-reader only,
lists the failed files and their reasons. The red line it replaces was
role="alert", so failures were announced; a hover tooltip is not.
- The reason now describes the remove button (aria-describedby on a
hidden span), the chip's only focusable control. Before, it was sr-only
text on a chip that can't take focus.
The composer showed each failed file's reason as a red line of text under
the chips. A failed file no longer blocks the send, since it is dropped
then, so that line was louder than it needed to be. DESIGN.md also says
status text of a sentence or more must not be a coloured paragraph.
The chip's warning icon already marks the failure. The reason is now a
Tooltip on the failed chip. It falls back to "Failed" when no reason is
known. It is also kept as sr-only text so screen readers still read it.
The native title moves from the chip to the truncating file name, where
DESIGN.md keeps title=, and is left off failed chips so the tooltip
doesn't double up.
Aborts /api/store_attachment, attaches a file, waits for the failed
chip's reason, then sends. Checks that one /stream request goes out with
no attachment ids, that the question is in the thread and the chip is
gone, and that the stored message has no attachments.
Uploads had no timeout, so a stalled XHR could leave a chip on
"uploading" forever, and a queued send waited with it. Network failures
also set no errorMessage, so the failed chip showed no reason.
guardUploadStall aborts an upload once it goes 120 s without progress.
It restarts on every upload progress event and again when the body
finishes sending. A large file on a slow link keeps going; only a silent
stall is cut off. That is why it is not xhr.timeout, which caps the whole
request against a 100 MB file limit. onerror, onabort and ontimeout now
mark the chip failed with "Upload failed. The file couldn't be read or
the connection dropped." (single and multi-file paths). A failed chip is
dropped at send, so a queued question always goes out in the end. The
5-minute processing cap is unchanged.
A queued send flushed by calling submitNow directly, which skipped the
loading check in canSubmit. submitNow clears the composer before the
consumer runs, and the consumer ignores a submit while status is
'loading'. If a retry or an edit was streaming when the attachments
settled, the question was cleared and never sent.
useArmedSend takes a canFlush flag. The composer passes whether it can
take a submit right now (not loading, not recording or transcribing).
The armed send stays queued until that holds, then goes out.
A failed attachment made getSendReadiness return 'blocked', so Send was a
no-op until the user found and removed the chip. On a phone the only hint
was a small red line after the tap, and a new user lost their question
that way. A failed file never recovers, so holding the send only costs
the question.
Readiness now ignores failed attachments: it waits only on uploading or
processing files. On submit the composer drops every failed attachment
and sends the text with the completed ones. The payloads already read
completed attachments only, so no failed id reaches the request or the
history. The hero-card path in Conversation also clears failed chips.
The failed chip still shows before Send and can be removed by hand.
Removes the send-blocked alert and its locale key.
arc53-machine