agent-chief MCP Server

How to use agent-chief MCP Server

uvx agent-chief demo        # zero keys, zero config, fully offline

{
  "$comment": "Register Chief as an MCP server (stdio). Claude Code: add under mcpServers in .mcp.json; Claude Desktop: claude_desktop_config.json. Tools exposed: propose, feedback, digest, policy, stats — see docs/protocol.md §2.",
  "mcpServers": {
    "chief": {
      "command": "uv",
      "args": ["run", "--with", "agent-chief", "python", "-m", "ingest.mcp_server"]
    }
  }
}

What is agent-chief MCP Server?

agent-chief is a Python MCP server with uvx/pip. Attention is your scarcest resource. Chief is the local-first layer that guards it — turning every agent, alert, and feed into one honest call: interrupt, or not. This profile combines repository metadata with install, compatibility, and usage signals so developers can quickly decide whether it fits their agent workflow before opening the source repository.

Key capabilities

• Tool access support
• Includes SKILL.md support
• Shell commands support
• Document generation support
• OpenAI provider support
• Anthropic provider support
• Ollama provider support

Technical details

Install or run agent-chief using uvx/pip. Check SmileLikeYe/agent-chief for the latest setup command.

• Built for Python
• Built for Rust
• Install or run with uvx/pip
• Uses stdio transport
• Uses SSE transport
• API key required
• Database access

Requirements and access

Security and permissions

Review permissions before connecting any MCP server to an agent. Pay special attention to whether it can read local files, write data, call external services, or perform destructive actions.

When to use agent-chief MCP Server

• Use it for connecting agents to external tools and data.
• Use it for integrations workflows.
• Use it for coding workflows.
• Use it for data workflows.

Built with

Explore related resources

Frequently asked questions