An MCP server for Claude Code that indexes past session transcripts and lets Claude search them semantically and by keyword. It returns raw conversation chunks and related metadata so Claude can reconstruct prior decisions, constraints, and failed attempts from older sessions.
From the registry: Pull Deep Memory from across your Claude Code Sessions — when you need it.

$ brew install ollamahttps://ollama.com$ bunx claude-find setupPlease install the `claude-find` MCP server into my current AI client (that's you).
Required prerequisites (do these first if not already done):
- **Install Bun** — Bun runtime is required to run the MCP server. Run: `brew install bun` (https://bun.sh/)
- **Install Ollama** — Ollama is required for embeddings. The setup command will start Ollama and pull the embedding model automatically. Run: `brew install ollama` (https://ollama.com)
- **Install Claude Code** — Claude Code must be installed because this server indexes and searches Claude Code session history.
- **Run initial setup** — Performs one-time setup: starts Ollama, pulls the embedding model, sets session retention to permanent, and registers the MCP server with Claude Code. Run: `bunx claude-find setup`
Canonical MCP server config (stdio transport):
- command: `bunx`
- args: ["claude-find"]
Note: The README does not provide a manual canonical MCP config snippet, but the server is registered via `bunx claude-find setup`. The simplest portable stdio invocation is `bunx claude-find`. After upgrading, `bunx claude-find index` can be run manually to rebuild the search index.
Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.MCP server for searching Airweave collections with natural language queries.