acp-router

Route plain-language requests to FoxFang ACP harness paths for session management.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/PotLock/foxfang --skill acp-router-potlock
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: acp-router
Source: https://github.com/PotLock/foxfang/tree/main/extensions/acpx/skills/acp-router
Command: npx skills add https://github.com/PotLock/foxfang --skill acp-router-potlock

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Routes plain-language requests for ACP harness environments (runtime or direct acpx) to the appropriate session path, unifying access and reducing manual orchestration.

Core Features & Use Cases

  • Route requests to the correct ACP harness path (runtime vs direct acpx) based on intent and availability.
  • Manage agent-id mappings and preflight checks for safe thread-spawn and lifecycle control across multiple agents (Pi, Claude, Codex, Gemini, etc.).
  • Support both session-based ACP runtime flows and direct acpx flows for coding tasks and harness experiments.

Quick Start

Try routing a simple coding task to the Codex ACP harness using a thread-based session.

Frequently Asked Questions about acp-router

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

FAQPage Schema
How do I route plain-language requests to the correct ACP harness path for session management?

ACP harness request routing directs plain-language intents to the appropriate runtime or direct acpx session path using preflight checks and agent-id mapping, ensuring deterministic task execution across multiple coding agents.

What is the difference between session-based ACP runtime flows and direct acpx flows?

Session-based ACP runtime flows manage continuous thread-spawn lifecycles, while direct acpx flows execute isolated coding tasks and harness experiments, with routing determined by intent and path availability.

How do I manage agent-id mapping for multiple coding agents like Pi, Claude, Codex, and Gemini?

Agent-id mapping assigns identity boundaries to each coding agent before preflight checks, enabling safe thread-spawn control and deterministic lifecycle management across Pi, Claude, Codex, and Gemini environments.

Can I use preflight checks to control thread-spawn across multiple coding agents?

Preflight checks validate harness path availability before thread-spawn, allowing graceful fallbacks when a runtime or direct acpx path is unavailable for the targeted coding agent session.

What happens when an ACP harness path is unavailable during task routing?

When an ACP harness path is unavailable, the routing mechanism triggers graceful fallbacks, preventing thread-spawn failures and ensuring safe, deterministic task routing without breaking the session lifecycle.

Do I need preflight checks to ensure deterministic task routing across coding agents?

Preflight checks are required to verify path availability and agent-id mapping before thread-spawn, guaranteeing deterministic task routing and safe session management across multiple coding agents.