Skip to content

SonAIengine/graph-tool-call

v0.6.1 Breaking

This release includes breaking changes for platform teams planning a safe upgrade.

Published 2mo MCP Developer Tools
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent agentic ai anthropic function-calling hybrid-search
+13 more
langchain llm mcp mcp-server openai openapi python swagger token-optimization tool-calling tool-retrieval workflow-engine zero-dependency

Summary

AI summary

Fixed wRRF weight bug so keyword_weight, graph_weight, and embedding_weight now apply correctly.

Full changelog

Bug Fixes

  • wRRF 가중치 버그 수정keyword_weight, graph_weight, embedding_weight가 실제로 적용되지 않고 모두 1.0으로 하드코딩되어 있던 버그 수정. embedding 활성화 시 설정값(keyword=0.2, graph=0.5, embedding=0.3)이 정상 반영됨
  • poetry.lock 동기화 — CI 실패 원인이던 lock 파일 갱신

New Features

  • Embedding 범용 어댑터 (wrap_embedding()) — LLM 어댑터(wrap_llm())와 동일한 패턴:
    • "openai/text-embedding-3-large" — OpenAI Embeddings API
    • "ollama/nomic-embed-text" — Ollama local embeddings
    • "sentence-transformers/all-MiniLM-L6-v2" — local sentence-transformers
    • "litellm/..." — litellm gateway
    • callable(list[str]) -> list[list[float]] — custom function
  • Corpus 기반 자동 stopword — DF 50% 이상 토큰을 자동 감지하여 BM25 query에서 제거. 하드코딩 제거, API마다 적응
  • set_weights() API — wRRF fusion 가중치를 사용자가 조절 가능:
    tg.set_weights(keyword=0.1, embedding=0.5)  # embedding 비중 증가
    

Search Quality Improvements

x2bee 1077 tools 기준:

  • "search products": ❌ getDeliveryMgmtList → ✅ getDisplayGoodsList (정확 매칭)
  • "회원 정보 수정": ⚠️ saveMemberDelivery → ✅ getMemberInfo (1위 개선)
  • "주문 취소": ✅ cancelOrder (유지)

322 tests passing, 6 skipped

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track SonAIengine/graph-tool-call

Get notified when new releases ship.

Sign up free

About SonAIengine/graph-tool-call

When tool count exceeds LLM context limits, accuracy collapses (248 tools → 12%). graph-tool-call builds a tool graph from OpenAPI/MCP specs and retrieves multi-step workflows via hybrid search (BM25 + graph traversal + embedding), recovering accuracy to 82% with 79% fewer tokens. Zero dependencies. Also works as an MCP Proxy — aggregate multiple MCP servers behind 3 meta-tools.

All releases →

Beta — feedback welcome: [email protected]