acp-router

Route coding requests to AI harnesses via ACP runtime or acpx sessions.

Updated Mar 10, 2026
One-click install
npx skills add https://github.com/lemonlqf/openclaw-rtsp --skill acp-router
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: acp-router
Source: https://github.com/lemonlqf/openclaw-rtsp/tree/main/extensions/acpx/skills/acp-router
Command: npx skills add https://github.com/lemonlqf/openclaw-rtsp --skill acp-router

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill streamlines the process of directing coding-related requests to specific AI coding environments like Pi, Claude Code, Codex, OpenCode, or Gemini, ensuring they are handled efficiently either through OpenClaw's ACP runtime or directly via acpx.

Core Features & Use Cases

  • Intelligent Routing: Automatically detects user intent to run code in various AI harnesses.
  • Flexible Execution: Supports both OpenClaw's ACP runtime for integrated sessions and direct acpx CLI execution for simpler relay tasks.
  • Use Case: When you ask OpenClaw to "run this Python script in Codex," this skill intercepts the request, identifies "Codex" as the target harness, and initiates the appropriate acpx command or sessions_spawn call to execute the script.

Quick Start

Ask OpenClaw to run a code snippet in your preferred AI coding harness.

Frequently Asked Questions about acp-router

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I route coding tasks to AI coding harnesses like Claude Code or Codex?

Routing coding tasks to AI harnesses like Claude Code or Codex is handled by detecting user intent and mapping the target agentId to execute the request via ACP runtime sessions or direct acpx CLI execution.

How does intent detection work when choosing between ACP runtime and direct acpx sessions?

Intent detection for ACP runtime versus direct acpx sessions works by evaluating the plain-language request, prioritizing the sessions_spawn tool for thread creation, and selecting the appropriate execution mode for the specified AI coding harness.

Can I use this to run a Python script in Gemini through OpenClaw ACP?

Yes, you can run scripts in Gemini through OpenClaw ACP by asking OpenClaw to execute the code, which triggers intent detection and initiates the appropriate sessions_spawn call or acpx command for the Gemini harness.

What happens if the OpenClaw ACP backend is unavailable during code execution?

If the OpenClaw ACP backend is unavailable, robust error handling and recovery policies trigger, managing ACP backend unavailability and acpx installation issues to ensure your coding-agent thread requests recover or fail gracefully.

Do I need to install acpx to route requests to AI coding assistants?

You need acpx installed for direct acpx-driven sessions, as the routing mechanism manages coding-agent thread requests by selecting between OpenClaw ACP runtime sessions and direct acpx CLI execution based on availability and installation status.

What is the best way to manage coding-agent thread requests across multiple AI environments?

The best way to manage coding-agent thread requests across multiple AI environments is using a routing system that prioritizes the sessions_spawn tool for thread creation and handles agentId mapping for various harnesses like Pi, Codex, and OpenCode.