Memnar

One source, every tool. Connect your Memnar once and any AI you open reads your memory live, always the current version. Memnar hosts your data in the EU and never sends it to a third party.

Reading your Memnar works on Plus and Pro. Saving new lessons back to Memnar during a chat requires a Business, Enterprise, or Edu plan.
- Copy the server URL below.
- In ChatGPT, go to Settings → Connectors → Create.
- Paste the URL. ChatGPT redirects you to your Memnar sign-in.
- Approve the connection.
- Test it: ask "What do you know about me?" and check that it responds as you.

The same connector works on claude.ai, Claude Desktop, and the Claude mobile app. A free Claude account allows one custom connector.

One click. Cursor installs the connector and writes the config file for you.
- Click Connect Cursor. Cursor opens with the Memnar server pre-filled.
- Click Approve in the dialog.
- Test it in a Cursor chat: ask "What do you know about me?"
-

claude mcp add --transport http memnar https://mcp.memnar.ai/mcp

Complete the sign-in in the browser that opens.

Test it: start a session and ask "What do you know about me?"

Requires Agent Mode (VS Code 1.102 or later).
-

In your project, create or open .vscode/mcp.json.

{ "servers": { "memnar": { "type": "http", "url": "https://mcp.memnar.ai/mcp" } } }

Test it: ask "What do you know about me?" and check that it responds as you.

Yes. The connection is read-only by default: Memnar sends your memory to the AI; the AI does not write back. The one exception is save-lesson: when you explicitly ask an AI to save something you learned in a chat, it adds that lesson to your Memnar.

Your AI-self files: identity, voice, and working rules, plus the patterns you have built up over sessions. The AI does not see your raw uploads. Memnar processes those when you build your profile; what comes through the connector is a structured summary in your own words.

In your AI tool's settings, remove the Memnar connector. On your Memnar Use page, under "Your connections," you can revoke access per tool at any time.

Reading your Memnar works on Plus and Pro. Saving a new lesson back to Memnar during a chat requires a Business, Enterprise, or Edu plan. This is a ChatGPT restriction, not a Memnar one.

Short notes on building a portable AI-self: new essays, product updates and behind the scenes thinking.

Local-first agent memory: a plain-Markdown Obsidian vault is the source of truth, with a rebuildable DuckDB index for hybrid BM25 + vector + graph recall.

Persistent memory and semantic search for AI coding assistants across sessions

Give your agent a memory: shared, cited, tenant-isolated knowledge-graph memory for any MCP host. Grounded answers from a local-first June endpoint — abstains rather than guesses.

Decentralized persistent memory for AI agents — encrypted vault storage built on Walrus and Sui.

Persistent memory for AI assistants and coding agents across ChatGPT, Claude, Cursor, and other MCP-compatible tools.

Your portable AI memory vault — memories, skills & configs, shared across every AI tool.

Local Work Model for AI agents that learns from real outcomes.

Adaptive MCP memory system for AI applications. Learns which retrieval strategies work for your data, scores results using cognitive science models, builds a knowledge graph automatically, and validates every parameter change against real query history before adopting it. Patent pending.

Auditable, self-improving knowledge & memory for AI agents over MCP — citation-enforced answers and a replayable why-trace, self-hosted on Postgres.