<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added canary build version support with automatic validation and
age-based restrictions for testing pre-release versions.
* **Chores**
* Enhanced Docker build process with multi-stage builds, image
optimization, and memory allocation improvements.
* Reorganized dependencies to distinguish development-only packages.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #14378** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a deprecated GET sign-out endpoint for backward compatibility
with legacy clients.
* **Improvements**
* Updated magic-link and OAuth flows to always generate and manage
client nonces; native clients use a nonce, web preserves cross-device
behavior.
* **Tests**
* Added tests covering the deprecated sign-out flow and OAuth preflight
client_nonce handling.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #14376** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Sign-out functionality now works in more scenarios, including when
headers are absent or duplicated.
* **Tests**
* Added test coverage for sign-out behavior across different header
configurations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #14373** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added client version tracking and validation to ensure application
compatibility across authentication flows and sessions.
* Enhanced OAuth authentication with improved version handling during
sign-in and refresh operations.
* **Bug Fixes**
* Improved payment callback URL handling with safer defaults for
redirect links.
* **Tests**
* Expanded test coverage for client version enforcement and session
management.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #14348** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a dedicated S3-compatible client package and expanded
S3-compatible storage config (endpoint, region, forcePathStyle,
requestTimeoutMs, minPartSize, presign options, sessionToken).
* Document sync now broadcasts batched/compressed doc updates for more
efficient real-time syncing.
* **Tests**
* New unit and benchmark tests for base64 utilities and S3 multipart
listing; updated storage-related tests to match new formats.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
# Summary
This PR fixes a server-side cookie parsing edge case where malformed
cookie values throw `URI malformed`, causing socket.io auth to fail and
clients to get stuck in infinite workspace loading/syncing.
# Observed Behavior
- User creates a cloud-backed workspace and invites another user to it.
- Second user accepts the invite, awaits approval, and attempts to load
the workspace, getting stuck in infinite loading state.
- `api/workspaces/<id>/docs/<id>` return 404 for those users, as the
workspace they are trying to access was not synced to the server.
- Server logs show socket.io `CONNECT_ERROR` with `URI malformed`, then
connection closed.
# Confirmed Trigger
An externally-managed `auth_session` cookie containing a raw `%` symbol
causes `decodeURIComponent` to throw. This matches the observed
socket.io `CONNECT_ERROR`, explaining why some users were affected while
the rest were not.
# Root Cause
The `parseCookies` function calls `decodeURIComponent` on every cookie
key/value without guard, so when a malformed percent-encoded value is
encountered, `decodeURIComponent` throws, which bubbles into the
socket.io auth middleware, aborting the connection.
# Fix
Wrap `decodeURIComponent` calls in `try/catch`, on failure falling back
to the raw key/value.
# Testing
- Manually regenerating the bad cookie until no malformed parts are
present resolves the issue.
- With the guard in place, affected users can open shared workspaces
with sync successfully completing.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved cookie parsing robustness so malformed cookie values no
longer cause errors; the system now preserves raw cookie values when
decoding fails.
* **Tests**
* Added test coverage to ensure cookie parsing handles invalid/malformed
cookie values without throwing.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #14251** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Current user profile now exposes access tokens, revealed tokens, and
detailed calendar accounts/subscriptions.
* Workspace now exposes permissions, calendars, calendar events, and a
workspace-scoped blob upload part URL.
* New document-update mutation for applying doc updates.
* **API Changes**
* validateAppConfig is now a query (mutation deprecated).
* Several legacy top-level calendar/blob endpoints deprecated in favor
of user/workspace fields.
* **Refactor**
* Calendar, blob-upload and access-token surfaces reorganized to use
user/workspace-centric fields.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #14203** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Removed deprecated database tables, enums and schema fields (cleanup
of legacy subscription, invoice, runtime settings and session expiry
data). This includes irreversible data removal for those legacy
elements.
* **Tests**
* Updated tests and test data to align with the cleaned-up schema and
removed fields.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Flexible blob uploads: GRAPHQL, presigned, and multipart flows with
per‑part URLs, abort/complete operations, presigned proxy endpoints, and
nightly cleanup of expired pending uploads.
* **API / Schema**
* GraphQL additions: new types, mutations, enum and error to manage
upload lifecycle (create, complete, abort, get part URL).
* **Database**
* New blob status enum and columns (status, upload_id); listing now
defaults to completed blobs.
* **Localization**
* Added user-facing message: "Blob is invalid."
* **Tests**
* Expanded unit and end‑to‑end coverage for upload flows, proxy
behavior, multipart and provider integrations.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a mutation to request/apply a subscription by transaction ID
(client mutation and server operation), returning subscription details.
* **Bug Fixes / Improvements**
* More robust external subscription sync with safer conflict detection,
optional short-lived confirmation, improved parsing and error logging.
* **Chores**
* Standardized time constants for clarity.
* **Tests**
* Updated subscription test data (expiration date) to reflect new
lifecycle expectations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
- **New Features**
- Safer, consistent file downloads with automatic attachment headers and
filenames.
- Smarter MIME detection for uploads (avatars, workspace blobs, Copilot
files/transcripts).
- Sensible default buffer limit when reading uploads.
- **Bug Fixes**
- Prevents risky content from rendering inline by forcing downloads and
adding no‑sniff protection.
- More accurate content types when original metadata is missing or
incorrect.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Copilot now defaults to the updated Claude Sonnet 4.5 model across
experiences for improved responses.
* **Chores**
* Consolidated available Anthropic models, removing older Sonnet 3.x
variants and standardizing Sonnet 4/4.5 options.
* Updated configuration defaults and schema mappings to reference the
new Sonnet 4.5 model.
* **Tests**
* Updated unit and end-to-end tests to reference the new model to ensure
consistent behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added an on-demand mutation to refresh the current user's
subscriptions, syncing with RevenueCat when applicable and handling
Stripe-only cases.
* Subscription variant normalization for clearer plan information and
consistent results.
* **Tests**
* Added tests for refresh behavior: empty state, RevenueCat-backed
multi-step sync, and Stripe-only scenarios.
* **Client**
* New client operation to invoke the refresh mutation and retrieve
updated subscription fields.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
- New Features
- Subscriptions now include an explicit "trial" flag so trialing users
are identified and treated correctly.
- Bug Fixes
- More robust handling when webhook fields are missing or null.
- Improved family-sharing detection to avoid incorrect async processing.
- Refactor
- Status determination and store resolution simplified to rely on
subscription data rather than event payloads.
- Tests
- Test fixtures updated to include trial and store details for accuracy.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* RevenueCat support: public webhook endpoint, webhook handler/service,
nightly reconciliation and per-user sync; subscriptions now expose
provider and iapStore; new user-facing error for App Store/Play-managed
subscriptions.
* **Chores**
* Multi-provider subscription schema (Provider, IapStore); Stripe
credentials moved into payment.stripe (top-level apiKey/webhookKey
deprecated); new payment.revenuecat config and defaults added.
* **Tests**
* Comprehensive RevenueCat integration test suite and snapshots.
* **Documentation**
* Admin config descriptions updated with deprecation guidance.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-419
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
- New Features
- New API to fetch available models for a prompt, returning default,
optional, and pro models with human‑readable names.
- Added temperature and topP settings to prompt configuration for finer
control.
- Refactor
- When no model is chosen, the default model is used instead of
auto-picking a pro model.
- Model metadata across providers now includes readable names, improving
listings and selection UX.
- Tests
- Updated test snapshots and descriptions to reflect the new
default-model behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
- New Features
- Added GPT-5 family and made GPT-5/-mini the new defaults for Copilot
scenarios and prompts.
- Bug Fixes
- Improved streaming chunk formats and reasoning/text semantics,
consistent attachment mediaType handling, and more reliable reranking
via log-prob handling.
- Refactor
- Unified maxOutputTokens usage; removed per-call step caps and migrated
several tools to a unified inputSchema shape.
- Chores
- Upgraded AI SDK dependencies and bumped an internal dependency
version.
- Tests
- Updated mocks and tests to reference GPT-5 variants and new stream
formats.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
- New Features
- Smarter embedding pipeline skips re-embedding when content hasn’t
changed; added content sanitization for embeddings and workspace content
retrieval.
- Bug Fixes
- Re-embedding now requires both a document update and the last
embedding being older than 10 minutes, reducing unnecessary work.
- Refactor
- Consolidated embedding preprocessing and moved sanitization utilities
into shared models; upserts now refresh stored content.
- Tests
- Expanded snapshot-based tests covering multiple time/age scenarios for
embedding decision logic.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #13434** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Style**
* Improved clarity of TypeScript error suppression comments across
various test files and helper scripts. Comments now specify the reasons
for ignoring specific type errors, enhancing code readability for
developers.
* **Chores**
* Updated inline comments without affecting application functionality or
user experience. No changes to features, logic, or test outcomes.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Updated scenario names and options for Copilot, including new and
renamed scenarios such as "audio_transcribing,"
"complex_text_generation," "quick_decision_making,"
"quick_text_generation," and "polish_and_summarize."
* Enhanced support for customizing and overriding default model
assignments in Copilot scenarios.
* **Bug Fixes**
* Improved consistency and clarity in scenario configuration and prompt
selection.
* **Documentation**
* Updated descriptions in configuration interfaces to better explain the
ability to use custom models and override defaults.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-404
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Introduced scenario-based configuration for copilot, allowing default
model assignments for various AI use cases.
* Added a new image generation model to the available options.
* **Improvements**
* Refined copilot provider settings by removing deprecated fallback
options and standardizing base URL configuration.
* Enhanced prompt management to support scenario-driven updates and
improved configuration handling.
* Updated admin and settings interfaces to support new scenario
configurations.
* **Bug Fixes**
* Removed deprecated or unused prompts and related references across
platforms for consistency.
* **Other**
* Improved test coverage and updated test assets to reflect prompt and
scenario changes.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Enhanced search functionality to include results from additional
"blob" data sources, providing more comprehensive search results.
* **Bug Fixes**
* Improved messaging to ensure "No results found" is only shown when no
relevant results exist across all data sources.
* **Tests**
* Updated test cases to reflect new keyword contexts, improving
validation accuracy for search-related features.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved permission checks when adding context blobs to ensure only
authorized users can perform this action.
* **Refactor**
* Streamlined background processing of blob embeddings by removing
user-specific parameters and simplifying job handling.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Updated test cases to use a new test asset describing AFFiNE.
* Adjusted assertions to check for "AFFiNE" in results instead of
previous keywords.
* Separated and refined the "Continue writing" test for clearer
validation.
* Improved assertion messages for clarity.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Introduced user access tokens, enabling users to generate, list, and
revoke personal access tokens via the GraphQL API.
* Added GraphQL mutations and queries for managing access tokens,
including token creation (with optional expiration), listing, and
revocation.
* Implemented authentication support for private API endpoints using
access tokens in addition to session cookies.
* **Bug Fixes**
* None.
* **Tests**
* Added comprehensive tests for access token creation, listing,
revocation, expiration handling, and authentication using tokens.
* **Chores**
* Updated backend models, schema, and database migrations to support
access token functionality.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added support for managing "blobs" in Copilot context, including
adding and removing blobs via new GraphQL mutations and UI fields.
* Introduced tracking and querying of blob embeddings within workspaces,
enabling search and similarity matching for blob content.
* Extended Copilot context and workspace APIs, schema, and UI to display
and manage blobs alongside existing documents and files.
* **Bug Fixes**
* Updated context and embedding status logic to handle blobs, ensuring
accurate status reporting and embedding management.
* **Tests**
* Added and updated test cases and snapshots to cover blob embedding
insertion, matching, and removal scenarios.
* **Documentation**
* Updated GraphQL schema and TypeScript types to reflect new
blob-related fields and mutations.
* **Chores**
* Refactored and cleaned up code to support new blob entity and
embedding logic, including renaming and updating internal methods and
types.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-357
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Improved prompt handling to conditionally include document fragments
based on the presence of documents in user queries.
* **Refactor**
* Updated system prompts to focus solely on document fragments, removing
references to file fragments for a more streamlined user experience.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Close [AI-396](https://linear.app/affine-design/issue/AI-396)
<img width="798" height="294" alt="截屏2025-07-25 11 30 32"
src="https://github.com/user-attachments/assets/6366dab2-688b-470b-8b24-29a2d50a38c9"
/>
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Summary by CodeRabbit
* **New Features**
* Introduced a "Section Edit" AI tool for expert editing of specific
markdown sections based on user instructions, preserving formatting and
style.
* Added a new interface and UI component for section editing, allowing
users to view, copy, insert, or save edited content directly from chat
interactions.
* **Improvements**
* Enhanced AI chat and tool rendering to support and display section
editing results.
* Updated chat input handling for improved draft management and message
sending order.
* **Other Changes**
* Registered the new section editing tool in the system for seamless
integration.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-392
fix AI-393
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Improved filtering of outdated document ID styles in embedding status
reporting, ensuring more accurate counts of embedded documents.
* Stricter rate limiting applied to workspace embedding status queries
for enhanced system reliability.
* **Bug Fixes**
* Resolved issues with duplicate or outdated document IDs affecting
embedding status totals.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Refactor
Align the Markdown conversion logic across all business modules：
1. frontend/backend apply: doc to markdown
2. insert/import markdown： use `markdownAdapter.toDoc`
> CLOSE AI-328 AI-379 AI-380
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Documentation**
* Clarified instructions and provided an explicit example for correct
list item formatting in the markdown editing tool.
* **Bug Fixes**
* Improved markdown parsing for lists, ensuring correct indentation and
handling of trailing newlines.
* Cleaned up markdown snapshot test files by removing redundant blank
lines for better readability.
* **Refactor**
* Updated markdown conversion logic to use a new parsing approach for
improved reliability and maintainability.
* Enhanced markdown generation method for document snapshots with
improved error handling.
* Refined markdown-to-snapshot conversion with more robust document
handling and snapshot extraction.
* **Chores**
* Added a new workspace dependency for enhanced markdown parsing
capabilities.
* Updated project references and workspace dependencies to include the
new markdown parsing package.
* **Tests**
* Temporarily disabled two markdown-related tests due to parse errors in
test mode.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-355
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added support for uploading a single file as an attachment when
creating chat messages, in addition to existing multiple file uploads.
* **Tests**
* Expanded test coverage to verify message creation with both single and
multiple file attachments.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-359
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added automated cleanup of embeddings for documents deleted or trashed
from workspaces.
* Introduced a new job to schedule and perform this cleanup per
workspace daily and on demand.
* Added new GraphQL mutation to manually trigger the cleanup process.
* Added the ability to list workspaces with flexible filtering and
selection options.
* **Improvements**
* Enhanced document status handling to more accurately reflect embedding
presence.
* Refined internal methods for managing and checking document
embeddings.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AI-367
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved document filtering to exclude settings documents and empty
blobs from embedding and status calculations.
* Enhanced embedding jobs to skip processing deprecated documents if a
newer version exists, ensuring only up-to-date documents are embedded.
* **New Features**
* Added a mutation to trigger the cron job for generating missing
titles.
* **Tests**
* Added test to verify exclusion of documents with empty content from
embedding.
* Updated embedding-related tests to toggle embedding state during
attachment upload under simulated network conditions.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
#### PR Dependency Tree
* **PR #13202** 👈
This tree was auto-generated by
[Charcoal](https://github.com/danerwilliams/charcoal)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Improved the "Transcript audio" text action by updating its default AI
model to "gemini-2.5-pro" for enhanced performance.
* Enhanced audio transcription accuracy by refining audio content
handling with a more specific audio format.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: DarkSky <darksky2048@gmail.com>
fix AI-350
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a new option to manually trigger the generation of missing
session titles via a GraphQL query.
* **Improvements**
* The process for generating missing session titles now considers all
eligible sessions, without limiting the number processed at a time.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
DarkSky