antiresonant/altRAG
Pointer-based skill retrieval for LLM agents. The alternative to vector DB RAG.
Builds a deterministic pointer table mapping document headings to exact byte offsets and line numbers, eliminating embeddings and vector databases entirely. The agent reads a lightweight skeleton (~100 tokens) during planning, then retrieves only the precise section needed—no similarity thresholds or fuzzy matching. Zero dependencies; integrates via auto-detected config files (`.cursorrules`, `CLAUDE.md`, `.clinerules`, etc.) and maintains freshness through git pre-commit hooks.
Available on PyPI.
Stars
4
Forks
—
Language
Python
License
MIT
Category
Last pushed
Mar 26, 2026
Monthly downloads
120
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/antiresonant/altRAG"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
neo4j/neo4j-graphrag-python
Neo4j GraphRAG for Python
FalkorDB/GraphRAG-SDK
Build fast and accurate GenAI apps with GraphRAG SDK at scale.
microsoft/graphrag
A modular graph-based Retrieval-Augmented Generation (RAG) system
Hawksight-AI/semantica
Semantica đŸ§ â€” A framework for building semantic layers, context graphs, and decision...
gusye1234/nano-graphrag
A simple, easy-to-hack GraphRAG implementation