V2 #21
Merged
V2 #21
Conversation
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
Pattern detection + quality scoring with zero dependencies Performance (300x faster than target): - 0.03ms per message (100 messages in 3ms) - 1.4ms average for real conversations - 100% detection rate on test patterns - Quality scoring <1ms Features: - Detects: decisions, problems/solutions, learnings, comparisons, anti-patterns - Quality scoring 0.0-1.0 with reasons - Auto-save threshold (0.7+), suggest threshold (0.4+) - Regex-based pattern matching, no NLP/ML - Super lightweight and fast Files: - src/context-extractor.ts: Pattern detection engine - src/context-quality-scorer.ts: Quality scoring system
Integrated context engine with MCP server for active collection New Tools: - conversation_checkpoint: Analyze conversation segments and auto-save valuable context - Auto-saves high confidence (>0.7) context automatically - Suggests medium confidence (0.4-0.7) for user review - Returns detailed analysis with scores and reasons - get_project_context: Enhanced to accept optional messages parameter - Analyzes messages on-the-fly - Shows detected patterns before saving - Provides save recommendations Features: - Real-time conversation analysis - Automatic context extraction (decisions, learnings, solutions) - Quality scoring with reasons - User-friendly output with emojis and formatting - Performance metrics included in response Integration: - ContextExtractor + ContextQualityScorer in server - Auto-save logic in handleConversationCheckpoint - Backwards compatible with existing save_decision tool This enables AI agents to pass conversation context for automatic capture, moving from passive (5% capture) to active (60-70% target).
Expanded storage with specialized tables for rich context capture New Database Tables: - learnings: Capture insights and discoveries - problem_solutions: Track problems and their solutions - comparisons: Store option comparisons with winners - anti_patterns: Record what NOT to do and why Schema Design: - Each table has confidence scores - Indexed by project_id and timestamp - Lightweight columns, fast inserts (<1ms) - Prepared statement caching for performance Storage Methods: - addLearning() / getLearnings() - addProblemSolution() / getProblemSolutions() - addComparison() / getComparisons() - addAntiPattern() / getAntiPatterns() Integration: - conversation_checkpoint now saves to proper tables - Each context type gets its own storage - Better querying and retrieval by type - Maintains backwards compatibility with decisions Performance: - All operations use prepared statements - Indexed for fast retrieval - Limit parameters prevent runaway queries This completes the active context collection architecture! Moving from 5% passive capture to 60-70% active capture with proper type separation and efficient storage.
…ols allowing developers to read, update, search memory simple. Remove todo-related schema, tools, and types; refactor workspace detection and Git hook manager tests
Owner
Author
|
@codex review |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1ffaf75bfe
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
…rror fix: remove BOM from setup script
…to-setup-tool feat: show auto-config status in setup wizard
…d schema migration
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Huge refactor,
Deprecated over 30 tools. Only kept 8 essential tools