Context Mem

Name: Context Mem
Author: JubaKitiashvili

by JubaKitiashvilicommunityv3.2.1

An MCP server that stores AI agent tool-call history in a local SQLite-backed memory system and syncs it to a browsable markdown vault. It also compresses observations with content-aware summarizers and supports hybrid retrieval across the stored memory.

From the registry: Persistent memory for AI agents — 98%+ retrieval recall, 99% token savings, 44 tools

GitHub Repository

Install & configure

Prerequisites

required
Install context-mem
Install the npm package globally or in your project, then run the init command to generate editor-specific MCP config files automatically.
```
$ npm i context-mem && npx context-mem init
```

Install in your client

Paste into the AI you're chatting with — it will figure out the rest

Please install the `context-mem` MCP server into my current AI client (that's you).

Required prerequisites (do these first if not already done):
- **Install context-mem** — Install the npm package globally or in your project, then run the init command to generate editor-specific MCP config files automatically. Run: `npm i context-mem && npx context-mem init`

Canonical MCP server config (stdio transport):
- command: `npx`
- args: ["context-mem","serve"]

Note: Manual MCP configuration uses stdio with `npx context-mem serve`. No API keys are required for core functionality. The README also mentions an optional local dashboard at `http://localhost:3141`, but that is not the MCP transport. Optional features mentioned outside MCP config include enabling the markdown vault in `.context-mem/config.json` and optional LLM reranking when `ai_curation.enabled = true`, but no MCP env vars are documented for these.

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.

Note: Manual MCP configuration uses stdio with `npx context-mem serve`. No API keys are required for core functionality. The README also mentions an optional local dashboard at `http://localhost:3141`, but that is not the MCP transport. Optional features mentioned outside MCP config include enabling the markdown vault in `.context-mem/config.json` and optional LLM reranking when `ai_curation.enabled = true`, but no MCP env vars are documented for these.