Agent♥︎Age
Catalog

CHAP Coordinator

Official

by BrightbeamAI · Python

Auditable records of human decisions over AI agent work. Approvals, edits, overrides, escalations.

io.github.BrightbeamAI/chap MCP Server

This MCP server adapter exposes a CHAP Coordinator as Model Context Protocol (MCP) tools. It wraps a Coordinator instance and publishes every CHAP method so an MCP client can drive a CHAP workspace directly. The project targets MCP 2026-07-28 and supports MCP 2025-11-25 clients as well (CHAP 0.2).

🛠️ Key Features

  • MCP server adapter for the CHAP Coordinator
  • Wraps a Coordinator instance
  • Exposes every CHAP method as an MCP tool
  • Tools compatible with MCP clients (e.g., Claude Desktop, Cursor, Claude Code)

🚀 Use Cases

  • Driving a CHAP workspace via an MCP client
  • Supporting auditable human decisions over AI agent work (approvals, edits, overrides, escalations)
  • Human-in-the-loop / human–AI collaboration workflows

⚡ Developer Benefits

  • Standardized integration through the MCP tool interface
  • Uses CHAP 0.2 conventions for coordinator-driven operations
  • Compatible with specified MCP client versions (2025-11-25, target 2026-07-28)

⚠️ Limitations

  • Requires Node 20+ at runtime
  • Relies on @brightbeamai/chap-coordinator and @modelcontextprotocol/sdk dependencies

Topics

human-ai-collaborationhuman-computer-interactionhuman-in-the-loophuman-robot-interactiona2aa2a-protocolai-agentsaudit-trailmcpmodel-context-protocolmulti-agent-systemspythonresponsible-aitypescriptai-governance-layerai-governance-tools