Orcho

•
4
•

Enables AI clients to control Orcho's durable AI software delivery lifecycle as a production harness and control plane.

Orcho is an MCP (Model Context Protocol) server that exposes the Orcho runtime to MCP-aware AI clients such as Claude Code, Cursor, and Zed. It functions as a robust control surface, allowing these clients to drive Orcho's agentic software delivery processes through a structured, typed interface. This eliminates the need for raw log scraping, enabling AI agents to programmatically act, observe, decide, inspect, and resume software delivery workflows. It facilitates a durable and automated AI-driven development lifecycle, ensuring seamless integration and efficient management of complex tasks and decisions within a production environment.

주요 기능

013 GitHub stars
02Programmatically start, resume, and cancel agentic software delivery runs with L4-test-pinned semantics.
03Support for AI agent decision-making on phase handoffs and post-release delivery or correction gates.
04Inspect run evidence, track changes with diffs, and measure resource consumption (tokens, duration, cost).
05Observe real-time run status, events, history, and tailored decision support for AI agents.
06Integrates with MCP-aware clients over stdio, providing a typed interface for control and observation.

사용 사례

01Empowering AI clients to directly manage and orchestrate automated software delivery pipelines.
02Providing a durable and inspectable control plane for agentic software development workflows.
03Integrating LLM-powered coding assistants with production software delivery and decision-making systems.