Feat/UI optimize 0702#243
Merged
Merged
Conversation
…conditions Extract buffer management from Team.tsx into a pure ConversationBufferManager class with a per-conversation phase state machine (idle/loading/ready/streaming). Key fix: DB load results during streaming phase write to cache only, never to display buffer, eliminating re-animation and message disappearance bugs. Also adds a loadingChat state so session switches show a spinner instead of the empty-chat greeting while messages load from DB. Includes 47 unit tests covering phase transitions, write routing, cache freshness, and race condition scenarios. Co-authored-by: Cursor <cursoragent@cursor.com>
- Unify DM and group chat A2A messaging via IM-style auto-reply model with processGroupChatReply orchestration and [NO_RESPONSE] termination - Add pending callback system for async message/tool result tracking - Enhance Notebook memory with Markdown format and structured sections - Strengthen DM prompt guidance to prevent acknowledgment ping-pong loops - Add server-side pure-acknowledgment auto-suppress for DM channels - Collapse agent-to-agent DM channels in Team Chat sidebar (collapsible) - Merge Mind tab into Overview tab in Agent Profile for consolidated view - Add /goal and /task slash commands in chat input - Update cognitive architecture, memory system, and API documentation Co-authored-by: Cursor <cursoragent@cursor.com>
- Add React.lazy code splitting for Team/Work/Settings/Store/Builder pages - Create shared dataCache layer with TTL + WS invalidation - Eliminate sidebar N+1 polling; use WS-driven last-message updates - Memoize AgentMessageBody/ChatTeamSidebar; cache segmentsToStreamEntries - Defer entity mention catalog loading until user types @ - Deduplicate useUnreadCounts to single module-level polling interval - Add LRU eviction (max 30) to sessionMsgCache in ConversationBufferManager - Gate BottomNav polling behind visibility check; increase interval to 30s - Reduce Home.tsx polling from 30s to 60s Co-authored-by: Cursor <cursoragent@cursor.com>
3-tier system prompt cache optimization: - Fix single-segment gate in Anthropic/OpenAI providers so triage cache hints work - Move scenario section to end of Tier 2 for better prefix stability across mode switches - Move colleague status and channel history to dynamic tier to protect Tier 2 cache - Add cache_control on Anthropic tool definitions - Split OpenAI system message by segments for implicit prefix caching - Collect cached_tokens metrics from OpenAI provider Tier 3 token reduction (~800-1800 tokens/call): - Remove available skills listing, replace with count hint in Identity - Remove project/shared deliverables injection (use deliverable_search on demand) - Remove task description from task board entries (use task_get for details) - Quantize notebook ages, mailbox elapsed time, and timestamp (minute-level) Co-authored-by: Cursor <cursoragent@cursor.com>
…completeness - Remove dead performTriage/triageJudge code path (~450 lines); only performDeliberation is used in production - Enforce strict completeness in complete_deliberation: tool returns error until every mailbox item is accounted for - Fix notify_user messages missing from agent context when user replies (stale in-memory session skipped DB reload) - Fix quoted/reply messages not visible to agent — prepend quoted text and persist replyTo metadata for chat_messages - Limit notebook agent entries from 15 to 4 - Fix pre-existing TS build error in anthropic.ts Co-authored-by: Cursor <cursoragent@cursor.com>
…ation Co-authored-by: Cursor <cursoragent@cursor.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.