GitHub - moorcheh-ai/memanto: Memory that AI Agents Love!

Memanto is a companion Memory Agent; an agent whose whole job is managing your other agents' memories.

⭐ Star the repo if Memanto is helping your agentic fleet.

It curates what's worth keeping, consolidates it across sessions, and briefs your agents the moment they need it, while you keep ownership of everything they learn. Works automatically with Claude Code, Cursor, Codex, and 20+ other agents. Fully convertible between semantic backend and Open Knowledge Format (*.md files in llm wiki style), so your memory estate is yours to inspect, export, and migrate anywhere — memanto migrate and it moves with you.

pip install memanto

Get started in 2 minutes

Works on macOS, Linux, and Windows.

Option A — Fully local (no account, no API key):

pip install memanto
memanto           # choose "On-Prem" — guides through Docker + Ollama setup

Requires Docker. Everything runs and stays on your machine.

Option B — Free cloud (no card, ~60 seconds):

pip install memanto
memanto           # choose "Cloud" — paste your free Moorcheh API key

Get your free API from : https://console.moorcheh.ai/api-keys

Switch between local and cloud at any time with memanto config backend.

What you get

Your fleet of agents start working

One memory, shared across the fleet. Your Cursor session doesn't know what Claude Code decided this morning. Your review agent doesn't know what your test agent already tried and rejected. Memanto gives them one shared memory; what any agent learns, every agent knows.

Every agent starts briefed. memanto agent bootstrap hands a new agent an intelligence snapshot of what the fleet already knows: the decisions, the constraints, the things that were tried and reversed. No warm-up prompt. No context-loading ritual before real work starts.

Scoped memory per agent. Each agent gets its own namespace, so your production-ops agent isn't reading your scratch experiments. Provision exactly what each one should know; nothing more.

The fleet gets smarter without you baby sitting it

Consolidation runs on schedule. memanto schedule runs daily: new memories curated, duplicates merged across agents, contradictions flagged for review. You come back to a fleet that knows more than it did yesterday, without having sorted anything yourself.

Contradictions surface instead of compounding. Two agents learn opposite things about the same system. memanto conflicts catches it, versions both, and brings it to you; rather than letting whichever wrote last silently win and propagate to everything downstream.

Memory that knows when. Query what the fleet knew last Tuesday (--as-of), or what changed since your last release (--changed-since). A preference from six months ago doesn't outrank yesterday's deadline.

See what your fleet knows

Full visibility over the memory estate. memanto status for registered agents, active sessions, and server health. memanto ui opens a local dashboard over everything the fleet has learned; browse it, search it, audit it.

Every memory carries reference and origin. Confidence score, source, timestamp, and what it superseded. When an agent acts on something, you can trace where that belief came from and when it entered the fleet.

Daily briefings on request. memanto daily-summary turns raw memory churn into a readable digest of what changed across your agents.

Nothing to provision, nothing to rewrite

One pip install. No vector DB, no embedding pipeline, no reranker, no schema migration, no backend to babysit. The retrieval engine ships in the box.

Works with the agents you already run. memanto connect claude-code — same for Cursor, Codex, Windsurf, Cline, Continue, Goose, Copilot and more. One command per agent. No code changes, no wrapper SDK, no rewrite of your agent loop.

Searchable the moment it's written. No LLM extraction at write time. No graph to rebuild. No indexing queue. remember returns and it's already retrievable — by every agent in the fleet.

It stays yours

Your fleet's memory is a file, not a hostage. memanto memory export --okf gives you plain Markdown — readable, diffable, committable. Switch stacks and take it with you. There's no lock-in because there's nothing to lock.

Runs entirely on your own machine. Local Docker + Ollama, no account, no API key, nothing leaves your infrastructure. Or free cloud, or your own hosting — switch any time with one command.

MIT licensed. No open-core bait, no feature gate, no rug pull.

Benchmarks

• 89.8% on LongMemEval and 87.1% on LoCoMo — outperforming Mem0, Zep, and Letta. Public datasets →
• Three primitives, not two: remember, recall, and answer LLM-grounded responses from memory, no extra API key.
• Single-query retrieval. No multi-stage pipelines, no graph schema, no rerankers.
• Typed semantic memory. 13 categories — instruction, fact, decision, goal, preference, relationship, and more.

See it in action

▶ Retrieving memory for an AI agentic workflow is much more than a search — 6:20 min

Setup & Demo

Local Dashboard For Best UX

Architecture

Memanto's retrieval is powered by Moorcheh, an information-theoretic semantic engine. It runs as a local Docker container (free, no account) or as a free cloud service (100K free operations) the memanto CLI manages either for you.

On-Prem

CLI Reference

For a complete command reference, see the CLI User Guide.

Supported Memory Types

instruction, fact, decision, goal, commitment, preference, relationship, context, event, learning, observation, artifact, error

Use memory types to categorize what you store so retrieval is cleaner and more controllable:

• Save with a specific type: memanto remember "User prefers concise answers" --type preference
• Filter by type when searching: memanto recall "user communication style" --type preference

📦 SDKs

• TypeScript / Node.js — @moorcheh-ai/memanto — boots a local Memanto server via uvx and exposes an ergonomic Memanto client (remember / recall / answer).

REST API

Memanto exposes a session-based REST API for programmatic access. Start the server locally:

Full endpoint reference is available at docs.memanto.ai/api and at http://localhost:8000/docs when the server is running.

Research

Memanto: Typed Semantic Memory with Information-Theoretic Retrieval for Long-Horizon Agents

@misc{abtahi2026memantotypedsemanticmemory,
      title={Memanto: Typed Semantic Memory with Information-Theoretic Retrieval for Long-Horizon Agents}, 
      author={Seyed Moein Abtahi and Rasa Rahnema and Hetkumar Patel and Neel Patel and Majid Fekri and Tara Khani},
      year={2026},
      eprint={2604.22085},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2604.22085}, 
}

Support

Have questions or feedback? We're here to help:

• Docs: https://docs.memanto.ai
• Discord: Join our Discord server
• Reddit: Join our Reddit community
• Email: support@moorcheh.ai
• X / Twitter: @moorcheh_ai

MIT License

English | Español | 简体中文 | 日本語