Skip to content
MCP server

Remembra Relay

Hands off a coding agent's session facts so the next agent can continue the work

Runs locallyStatus and history
Install
Pricing
Freemium
Platforms
CLI, API, self-hosted
Runs
Locally
Version
0.17.8
Listing updated
3 days ago

About Remembra Relay

Remembra Relay is a freemium MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.

Handoffs are built from facts taken from git and the session transcript, such as commits on the branch, files changed and test results, plus the open item left behind. When one agent stops because it ran out of credits or hit a usage limit, the next agent reads a short brief of about 1,500 tokens and carries on, even on another machine or from another vendor.

Session hooks write and read the handoffs automatically, and the remembra-install command shows what it will add to each agent it finds before writing anything. Agents can also call the session_brief tool directly, and a handoff written through close_session is labelled as declared by the agent and not checked.

The service is a hosted cloud memory with a free key, and the README and site describe a self-hosted option. An API key is required, and the agent id setting lets you see which tool wrote each handoff.

Install

Run in a terminal

claude mcp add remembra --env REMEMBRA_API_KEY='<REMEMBRA_API_KEY>' --env REMEMBRA_URL='<REMEMBRA_URL>' -- uvx remembra-mcp

Add to claude_desktop_config.json

{
  "mcpServers": {
    "remembra": {
      "command": "uvx",
      "args": [
        "remembra-mcp"
      ],
      "env": {
        "REMEMBRA_API_KEY": "<REMEMBRA_API_KEY>",
        "REMEMBRA_URL": "<REMEMBRA_URL>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "remembra": {
      "command": "uvx",
      "args": [
        "remembra-mcp"
      ],
      "env": {
        "REMEMBRA_API_KEY": "<REMEMBRA_API_KEY>",
        "REMEMBRA_URL": "<REMEMBRA_URL>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "remembra": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "remembra-mcp"
      ],
      "env": {
        "REMEMBRA_API_KEY": "<REMEMBRA_API_KEY>",
        "REMEMBRA_URL": "<REMEMBRA_URL>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "remembra": {
      "command": "uvx",
      "args": [
        "remembra-mcp"
      ],
      "env": {
        "REMEMBRA_API_KEY": "<REMEMBRA_API_KEY>",
        "REMEMBRA_URL": "<REMEMBRA_URL>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "remembra": {
      "command": "uvx",
      "args": [
        "remembra-mcp"
      ],
      "env": {
        "REMEMBRA_API_KEY": "<REMEMBRA_API_KEY>",
        "REMEMBRA_URL": "<REMEMBRA_URL>"
      }
    }
  }
}

Configuration

  • REMEMBRA_API_KEYRemembra API key (dashboard: Settings > API keys)Required
  • REMEMBRA_URLRemembra server URL: https://api.remembra.dev, or your self-hosted serverRequired
  • REMEMBRA_AGENT_IDName this agent's handoffs are recorded under, e.g. claude-code, codex, cursor

Tools this server provides (2)

Taken from the project documentation; the running server may differ.

session_brief
Returns a short text brief of who worked last, what is done, what is failing and the next step.
close_session
Writes a handoff holding the facts the agent declares, labelled as not checked.

What it does

Handoffs from real facts
Records commits, changed files and test results from git and the session transcript.
Session hooks
Writes a handoff when a session ends and reads the brief when the next one starts.
Capped session brief
The brief is limited to about 1,500 tokens so it fits at the start of a session.
Cross-vendor pickup
A different agent on another machine can continue from the same trail.
Reviewable install
remembra-install shows what it will add to each agent and writes only after you agree.

Good for

  • Continue a task in a second coding agent after the first hits its usage limit
  • Move work between a laptop and a desktop without retyping context
  • Keep an audit trail of which agent did what on a project
  • Start a new day from the last agent's open items instead of a blank prompt

Pricing

Freemium, from $12 a month

Free plan with 500 credits a month; Solo $12/mo or $120/yr

Prices change. Check remembra.dev before you buy.

Remembra Relay FAQ

What happens when an agent runs out of credits?

You end its session and the next agent picks up from the handoff left by the session hook, on the same or another machine.

Does it need an API key?

Yes. The package asks for REMEMBRA_API_KEY, and the server URL defaults to the hosted API or can point at a self-hosted server.

How are handoffs labelled?

Hook-written handoffs use facts from git and the transcript; ones written through close_session are marked as declared by the agent and not checked.

How much does it cost?

There is a free plan with 500 credits a month and 3 projects. Paid Solo is listed at $12 a month, or $10 a month billed yearly.

Is Remembra Relay free?

Remembra Relay has a free plan, with paid plans for more usage or features. Paid plans start at $12 a month.

How do I add Remembra Relay to Claude Code?

Run this command in your terminal: claude mcp add remembra --env REMEMBRA_API_KEY='<REMEMBRA_API_KEY>' --env REMEMBRA_URL='<REMEMBRA_URL>' -- uvx remembra-mcp

What are the best alternatives to Remembra Relay?

The closest alternatives to Remembra Relay are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Remembra Relay

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools