{"repo":"Anna-Team/AnnaDB","free":true,"listed":false,"github":"https://github.com/Anna-Team/AnnaDB","clone":"git clone https://github.com/Anna-Team/AnnaDB.git","description":"Developer-first database","language":"Rust","stars":62,"topics":["database","nosql"],"license":null,"category":"databases-storage","readme_excerpt":"AnnaDB — Embedded AI Agent Memory Engine AnnaDB is an embedded, local-first memory engine for AI agents. Document-oriented, link-based graph model with vector search, zero-cycle graph traversal, and HTTP-native protocol. Purpose-built for agents that need semantic recall, structured relationships, and durable memory. Quick start Features Feature Description --------- ------------- Vector search HNSW index, cosine/euclidean/dot similarity, optional OpenAI embeddings Graph operations Typed edges, neighbors, BFS traversal, shortest path, ego graph Zero config Keyword recall works immediately; vector search via env var Embedded mode SQLite-like open() API for Rust + PyO3 Python bindings HTTP server POST /tx with TySON, GET /health, zero native dependencies Persistence WAL + periodic snapshots, crash recovery BSL licensed Free for individuals, academics, non-profits, companies < $5M Architecture Usage Embedded (Rust) HTTP server With vector search opencode Integration AnnaDB serves as persistent memory for opencode coding sessions. The agent remembers decisions, files, bugs, and conventions across sessions. The skill teaches the agent to use these tools: Tool TySON Purpose ------ ------- --------- memory remember remember s collection ... s content ... Store facts, decisions memory recall recall s collection ... s query ... Retrieve relevant context memory relate relate s from ... s to ... Connect memories memory forget forget s link ... Delete memories memory inspect list collect","default_branch":null,"files":null,"tree":[],"storefront":"/r/Anna-Team","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/Anna-Team/AnnaDB/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}