Conversation
…torless guide Replace basic project documentation with detailed Vectorless Architecture Guide covering overview, architecture components, indexing pipeline, retrieval engine features, and performance characteristics. The new content explains how Vectorless uses tree-based navigation instead of vector embeddings for document retrieval through LLM-powered reasoning.
… information - Add detailed error information to the query failure error message - Include source and error details for each failed document in the error message - Format error details as "source (error)" pairs joined by semicolons
Add support for optional reasoning index in document retrieval pipeline. The reasoning index enables fast-path lookups for specific query patterns like document summaries and overview questions. Update the Engine and Retriever interfaces to pass reasoning index through the query flow. refactor(pilot): update guide_start method signature Change the guide_start method in Pilot trait to accept a start_node parameter, allowing pilots to evaluate the correct children instead of always using root's children. Update LlmPilot and NoopPilot implementations accordingly. feat(toc-navigator): implement keyword overlap fallback for abstract queries Add Phase C to ToCNavigator that uses keyword overlap matching against section summaries when BM25 produces no results. This handles abstract queries like "What is this project about?" where query terms don't appear in section titles but summaries contain relevant matches. fix(evaluate-stage): use BFS for leaf content collection Switch from DFS to BFS (FIFO) traversal in collect_leaf_content to preserve document order - ensuring the first section in the document appears first in the output. feat(evaluate-stage): detect stagnant candidate loops Implement stagnant candidate detection in EvaluateStage to identify when candidates remain unchanged between iterations, indicating that further backtracking won't improve results. refactor(search-stage): enhance summary query detection Improve summary query detection logic to handle phrase patterns with filler word removal, making it more effective at identifying overview requests like "what is this project about". refactor(persistence): fix serialization drift issues Change PersistedWrapper to store payload as serde_json::Value to avoid HashMap key ordering drift during repeated serialization/deserialization cycles.
…ndling - Change return type from `Result<(), Box<dyn std::error::Error>>` to `vectorless::Result<()>` - Rename client variable to engine for consistency - Update workspace path to "./worksspace_flow_example" - Fix typo in API key format from "sk-..." to "sk..." - Add proper endpoint configuration with "https://api" - Simplify error mapping by removing explicit type annotation - Update query text to "What is the seconds for complex multi-hop?" - Fix content preview truncation logic - Comment out document removal step for cleanup
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
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
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.