Skip to content
MCP server

sealkeep

Let AI agents run commands with your secrets while the values stay hidden and redacted

Runs locallyStatus and history
Pricing
Open source
Platforms
CLI, macOS, Linux
Runs
Locally
Version
0.5.4
Listing updated
2 days ago

About sealkeep

sealkeep is an open-source MCP server for security. It runs locally on your machine and is started by your AI client.

You store each secret once under a path-like name such as shared/stripe/test/SECRET_KEY, held in the operating system keyring (macOS Keychain, Secret Service or Windows Credential Manager) with an optional Vault KV v2 replica. The agent finds names with a list command, then asks sealkeep to run a command with a secret. The value goes into that one command's environment and is replaced by a placeholder like [sealkeep:NAME] in all output, so it never enters the transcript, context or a file.

The MCP server exposes three tools: list_secrets, run_with_secrets and store_status. None returns a value. A separate guard hook for Claude Code, Codex and Cursor refuses tool calls that would print a secret, such as cat .env, kubectl get secret or vault kv get. Revealing a value works only from a real terminal.

The CLI also imports dotenv and INI files, scans a machine for .env files without printing values, supports aliases so one value serves several projects, and can write a temporary dotenv file for tools that require one. An install script and a Claude Code plugin set up the server, skill and hook.

Install

Run in a terminal

claude mcp add sealkeep -- npx -y @munhq/sealkeep

Add to claude_desktop_config.json

{
  "mcpServers": {
    "sealkeep": {
      "command": "npx",
      "args": [
        "-y",
        "@munhq/sealkeep"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "sealkeep": {
      "command": "npx",
      "args": [
        "-y",
        "@munhq/sealkeep"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "sealkeep": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@munhq/sealkeep"
      ]
    }
  }
}

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

{
  "mcpServers": {
    "sealkeep": {
      "command": "npx",
      "args": [
        "-y",
        "@munhq/sealkeep"
      ]
    }
  }
}

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

{
  "mcpServers": {
    "sealkeep": {
      "command": "npx",
      "args": [
        "-y",
        "@munhq/sealkeep"
      ]
    }
  }
}

Configuration

  • SEALKEEP_CONFIGPath of the config file (stores, aliases, guard rules). Default: the config folder of the OS. With no file, sealkeep uses the OS keyring.

Tools this server provides (3)

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

list_secrets
List stored secret names
run_with_secrets
Run a command with named secrets injected and output redacted
store_status
Show the status of the secret stores

What it does

Hidden secret injection
Values enter one command's environment only.
Output redaction
Values become placeholders in stdout and stderr.
Guard hook
Blocks commands that would print a secret.
Secret scan and import
Finds .env files and imports them without showing values.
Aliases
One stored value shared across projects.
Keyring storage
Uses the OS keyring, with optional Vault replica.
Client installer
Sets up Claude Code, Codex and Cursor.

Good for

  • Let an agent call an API with a key it cannot read
  • Run a dev server with environment variables from a folder of secrets
  • Give a browser-testing tool a test login without exposing the password
  • Find stray .env files on a machine

Pricing

Open source

Free to use; license not published

sealkeep FAQ

Can the agent read the secret value?

No. No MCP tool returns a value, and output is redacted to a placeholder.

Where are secrets stored?

In the OS keyring, with an optional Vault KV v2 replica.

What does the guard hook do?

It refuses tool calls that would print secrets, for example cat .env or sealkeep get.

How do I install it?

Run npx -y @munhq/sealkeep as an MCP server, or use the Claude Code plugin or install script.

Is sealkeep free?

Yes. sealkeep is open source, so you can use and self-host it for free.

How do I add sealkeep to Claude Code?

Run this command in your terminal: claude mcp add sealkeep -- npx -y @munhq/sealkeep

What are the best alternatives to sealkeep?

The closest alternatives to sealkeep are REA, Agent Roast, Midplane, Leash and Tinplate Edge AI Safety, Security & EU Compliance.

Alternatives to sealkeep

  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. Agent Roast

    MCP server

    Offline review that flags agent changes which weaken tests, assertions or CI

    Runs locally
    • Open source
    • 4 Hacker News points
    • Developer tools
  3. Midplane

    MCP server

    Postgres gateway for AI agents with table grants, column masking and approval for writes

    Runs locally
    • Open source
    • 5 GitHub stars
    • Databases
  4. Leash

    MCP server

    Review what your AI touched, approve risky actions and undo supported changes

    • 1 tool
    • Remote
    • Free
    • 1 GitHub star
    • Security
  5. SecureCloud

    MCP server

    SentinelOne and IT security catalog, prices, quotes and checkout links for Israel

    • 7 tools
    • Remote
    • Free
    • Security