Skip to content

Latest commit

 

History

History
41 lines (33 loc) · 1.67 KB

File metadata and controls

41 lines (33 loc) · 1.67 KB

Roadmap

Recently Completed

  • Parser modules extracted from stages.py into core/parsers/ package (python.py, javascript.py). stages.py is now thin orchestration only.
  • SQLite FTS5 search index (search.db) built by load, queried by search. Falls back gracefully to O(n) JSONL scan before first load.
  • analyze hotspots and analyze metrics fixed to use CALLS_FUNCTION edges for accurate fan-in/fan-out coupling metrics.
  • start-neo4j.ps1 Docker helper script. Reads credentials from .env, creates or restarts neo4j-cgx, polls until bolt is ready.
  • AST nested function scope handling fixed to match tree-sitter behavior.
  • _looks_write_query false-positive guard replaced with regex clause boundaries.
  • _resolve_query restricted to .cypher files (closed arbitrary file read).

Next Milestones

  1. Data Quality

    • Expand parser coverage for class and method constructs in JS/TS.
    • Add fixture sets for mixed-language monorepos.
    • Add TypeScript-specific type annotation extraction.
  2. Analysis Depth

    • Add path-sensitive impact scoring.
    • Add churn-aware hotspot ranking from VCS metadata.
    • Wire ask command to a real LLM backend.
  3. Scale

    • Add chunked parallel extract/load execution.
    • Add memory-bound streaming for large repos.
  4. Developer Experience

    • Add codegraphx pipeline run orchestration command.
    • Add machine-readable run manifest for every stage.
    • Add shell completion generation for PowerShell and bash.
  5. Governance

    • Expand threat model and security test fixtures.
    • Add compatibility matrix documentation for Python/OS combinations.
    • Add integration test suite for Neo4j-backed commands.