memsearch

memsearch is a cross-platform semantic memory layer for AI coding agents backed by Markdown and Milvus.

MemSearch provides cross-platform semantic memory for AI coding agents. It captures conversations automatically and supplies semantic recall with zero configuration through plugins for Claude Code, Codex, DeepSeek use, OpenClaw, and OpenCode. Markdown files serve as the source of truth, while Milvus acts as a shadow index. It features a 3-layer recall system using dense vectors, sparse BM25, and RRF reranking, alongside optional background tasks to maintain PROJECT.md and USER.md notes and distill repeatable workflows into installable agent skills. Developers can also use the CLI and Python API to build memory into custom agents.

memsearch at a glance

Best for: Software engineer
Runs on: CLI, API, self-hosted, open source
Pricing: Free tier available via Zilliz Cloud; open source code available.
Released: February 2026
Category: AI agent tool
Website: github.com/zilliztech/memsearch

Key features

Automatic conversation capture
3-layer progressive recall
Dense vector and sparse BM25 search
Background project and user note maintenance
Procedural skill distillation from sessions
Markdown-based durable storage
CLI and Python API support

Use it for

Resume debugging threads
Recover decision rationale
Trace feature history
Do code archaeology
Carry context across agents

Companions

Tools that articles about memsearch recommend using with it, each with its job.

Claude Code — Built to work with Claude Code.
Codex — Built to work with Codex.
Context7 — Provides up-to-date documentation for AI agents and code editors.
Superpowers — Brainstorm, plan and execute skills for bigger tasks.
Playwright MCP — Enables LLMs to interact with web pages using structured accessibility snapshots.
GitHub MCP Server — Reads pull requests, issues and code search on GitHub.

Alternatives to memsearch

claude-mem — It shares persistent, searchable agent memory across sessions and devices.
Memvid — Single-file architecture replaces complex RAG pipelines with portable storage.
memU — It gives every connected agent a shared and inspectable memory layer.
Supermemory — Integrates long-term memory into AI coding agents with a single prompt.
Mem0 — It provides drop-in persistent memory for AI agents without changing pipelines.
MemPalace — Achieves high recall with local-first storage and zero API calls.
AgentMemory — It runs locally with zero external databases and achieves high retrieval recall.
AI Memory — Shared memory for coding agents that works across machines and team members.
MuscleMemory — It cuts input tokens, tool calls, and execution time while maintaining task success.
ProjectMem — Records project history locally to prevent AI from repeating failed work.
Engram — Single Go binary with zero dependencies and built-in full-text search.
OptMem — Provides permanent agent memory with fast lookups and no background processes.

All memsearch alternatives