{
  "markdown": "# memfleet\n\nShared, governed memory for fleets of AI agents. Contributions pass\njudgment; every entry carries provenance; operators see everything.\n\nThis repository is the server home for the **memfleet MCP server** —\n`com.memfleet/memfleet` on the\n[official MCP Registry](https://registry.modelcontextprotocol.io). The\nplatform lives at [memfleet.com](https://memfleet.com); the command-line\nclient ships as [`memfleet` on PyPI](https://pypi.org/project/memfleet/).\n\n## What memfleet is\n\nmemfleet is the hosted platform for\n[Strata](https://github.com/oren198/Strata), the open-source memory engine.\nYou create a **Workspace** (a full memory fleet), register agents that\nreceive keys, and operate everything from a web Console while your agents\nread and contribute over MCP or REST. Judgment runs as a platform service.\n\nAgent teams forget: every session starts from zero and lessons learned by\none agent never reach the rest. With memfleet, agents send\n**contributions**; nothing writes to shared memory directly. A **judgment**\nstep decides what enters and whether it lands as a binding **directive** or\ninformative **context**. Memory is organized in **scopes** arranged in\n**strata**, each agent reads a composed **perspective** of what its scope\ncan see, and the **record** keeps provenance for every entry. Operators\nmanage the whole lifecycle from a web Console, including seeing any agent's\nexact perspective.\n\n## The MCP server\n\n- **Endpoint**: `https://memfleet.com/mcp` (streamable HTTP)\n- **Auth**: `Authorization: Bearer <agent key>` — minted for a Registered\n  Agent in your Workspace's Console; resolve it from an environment\n  variable, never commit it.\n- **Tools**: read perspective, list scopes, read scope summary and record,\n  contribute, publish, withdraw, session stats, session closeout.\n- **REST**: the same operations at `/agent/v1` for anything that speaks\n  HTTP.\n\nAgents read and contribute over MCP or REST, so any MCP client can join\nthe fleet; first-class setup ships for Claude Code today.\n\nThe [`server.json`](./server.json) in this repository is the registry\nlisting manifest for `com.memfleet/memfleet`.\n\n## Quickstart\n\n```\npipx install memfleet\nmemfleet connect                                        # once per machine — one browser approval\nmemfleet run --profile 'Fleet/backend-bot' claude       # per project / per session\n```\n\n`connect` links the **machine** to your account. `run` binds the\n**session** to a **Registered Agent** — creating it if it does not exist\nyet — writes the project wiring, and launches the command. The setup it\nwrites is secret-free: no key is ever written to disk in plaintext.\n\n## What the fleet gets\n\n- **Governed memory, not a shared scratchpad.** Every contribution passes\n  judgment before it enters shared memory; every entry lives in a scope and\n  carries provenance in the record.\n- **Memory that doesn't eat your context window.** An agent reads its\n  perspective: a compact, composed summary of what its scope can see, not a\n  transcript dump. That keeps the tokens your agent spends on memory small,\n  and reading a perspective is never metered.\n- **The operator is sovereign.** See exactly what any scope's agents see,\n  and steer memory directly. No hidden state.\n- **The engine is open source.** Run\n  [Strata](https://github.com/oren198/Strata) yourself, or let memfleet run\n  the fleet for you.\n\n## Links\n\n- Platform and Console: [memfleet.com](https://memfleet.com)\n- Pricing: [memfleet.com/pricing](https://memfleet.com/pricing)\n- Client on PyPI: [pypi.org/project/memfleet](https://pypi.org/project/memfleet/)\n- Engine, open source: [github.com/oren198/Strata](https://github.com/oren198/Strata)\n\n## License\n\n[MIT](./LICENSE)\n",
  "bytes": 3714,
  "sha": "2ce2f33bfe11980769ed803d04d8c35f1b16c0c847f419c48474f233726afea0",
  "repo_slug": "oren198/memfleet",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_com_memfleet_memfleet_696db279/readme"
}