Touchstone — know if you broke your RAG. An MCP server and CLI that score retrieval quality against a golden set: recall@k, MRR and hit rate, with no LLM calls.

claude-code embeddings evaluation evaluation-metrics llm mcp model-context-protocol openai pgvector postgres python rag retrieval retrieval-augmented-generation vector-search
2 Open Issues Need Help Last updated: Sep 1, 2026

Open Issues Need Help

View All on GitHub
enhancement good first issue

Touchstone — know if you broke your RAG. An MCP server and CLI that score retrieval quality against a golden set: recall@k, MRR and hit rate, with no LLM calls.

Python
#claude-code#embeddings#evaluation#evaluation-metrics#llm#mcp#model-context-protocol#openai#pgvector#postgres#python#rag#retrieval#retrieval-augmented-generation#vector-search
enhancement help wanted

Touchstone — know if you broke your RAG. An MCP server and CLI that score retrieval quality against a golden set: recall@k, MRR and hit rate, with no LLM calls.

Python
#claude-code#embeddings#evaluation#evaluation-metrics#llm#mcp#model-context-protocol#openai#pgvector#postgres#python#rag#retrieval#retrieval-augmented-generation#vector-search