Memora: Persistent Memory Layer for AI Agents via MCP and Semantic Search
tom_doerr · x · 2026-08-07
Memora is an open-source project designed to provide persistent memory capabilities for AI agents.
- Core Features: As an MCP (Model Context Protocol) memory layer, it offers structured storage, semantic retrieval, graph relations, and source-backed cross-session context.
- Workflow: It absorbs agent workflows into a durable graph memory, which can then be queried using the memorydigest(topic) function to retrieve relevant memories, TODOs, related edges, and source IDs.
- Storage & Sync: Built on SQLite with optional cloud synchronization support (S3, R2, D1).
- Extensibility: Includes features for document handling, LLM deduplication, linking, and Neovim integration.
More from coding & agent
- Open-Sourcing Mandate: A Financial Stack Giving AI Agents Economic Autonomy — RichardsonDx · 2026-08-07
- Pydantic Creator Prefers Claude Code, Leaves OpenAI Free Tokens Unused Due to Poor UX — samuelcolvin · 2026-08-07
- Mediabunny optimizes large file streaming with on-demand downloads and memory control — Vjeux · 2026-08-07
- The Future is AI-to-AI Loops: Agents Will Negotiate APIs and Report Bugs — DanielLockyer · 2026-08-07
- Developers Frustrated by Claude Code Harness, Leaving Free Tokens Unused — intellectronica · 2026-08-07
- AI Engineering is Shifting to Systems Design: Context and Routing Take Center Stage — Deep_Ladder_4679 · 2026-08-07