HN Showcase
Back to browse
GitHub Repository

Federated, local-first search for an AI — one query across transcripts, files, knowledge graph, vector store, and the web, fused by trust-weighted RRF. Apache-2.0.

18 starsPython

Magpie-Search

by Floukie·Jun 23, 2026·2 points·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidBig BrainNiche Gem

Trust-tier ranking beats naive RAG fusion when sources conflict.

Strengths
  • •Trust tiers (fact > reference > lead > stale) add verifiable source credibility.
  • •SQLite FTS5 + sqlite-vec enables hybrid keyword and semantic search locally.
  • •MCP integration means agents access all six sources through one protocol.
Weaknesses
  • •Agent memory and retrieval is crowded (Cursor, Continue, LangChain already here).
  • •18 stars suggests early stage; production reliability unproven at scale.
Category
AI/ML
Target Audience

AI developers building agent systems with memory and retrieval needs

Similar To

LangChain · LlamaIndex · Mem0

Similar Projects

Data●●Solid

Vector databases are the wrong primitive for AI agents

Knowledge graphs beat vector similarity for structured relationship queries.

Big BrainBold Bet
ajainvivek
113mo ago
AI/ML●●Solid

AgentKV – SQLite for AI agent memory (MMAP vector+graph DB)

Single-file mmap storage plus an HNSW vector index and explicit graph edges is an elegant, practical combo — think "SQLite for agent memory" with CRC-32 crash recovery and zero-server convenience. The C++20 core + nanobind gives zero-copy NumPy views and GIL-free searches, and the claimed FAISS-like throughput makes this genuinely interesting for local setups; main caveat is build/toolchain friction and how rich the surrounding ecosystem becomes.

WizardryNiche Gem
shiwang_khera
104mo ago