Buckeyes22

Subagent Model Routing — AI skill for Claude Code

AI community

Route Claude Code, Codex, and Copilot work across installed model CLIs with delegation guardrails and an observation-driven routing ledger.

How to install Subagent Model Routing

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Buckeyes22/subagent-model-routing and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Subagent Model Routing does

Route Claude Code, Codex, and Copilot work across installed model CLIs with delegation guardrails and an observation-driven routing ledger.

Alternatives in AI

Claude Task Master — by eyaltoledano - A task management system for AI-driven development with Claude, designed to work seamlessly 26k ★
Embeddedskills — An open-source collection of embedded development and debugging skills for Claude Code, Copilot, TRAE, and oth 593 ★
HarnessKit — More than a skill manager — manage skills, MCP servers, plugins, hooks, CLIs, configs, memory & rules across e 416 ★

README

subagent-model-routing

[!IMPORTANT] This repository is archived. Agent Routing development, documentation, and releases have moved to [Pitwall](https://github.com/Buckeyes22/pitwall-gpu-broker). Install the current component from the namespaced [Agent Routing v0.10.0 release](https://github.com/Buckeyes22/pitwall-gpu-broker/releases/tag/agent-routing%2Fv0.10.0) or use the [Pitwall Agent Routing documentation](https://github.com/Buckeyes22/pitwall-gpu-broker/tree/main/docs/agent-routing). The material below documents the historical standalone releases through `v0.7.1` and is no longer maintained.

You probably pay for more than one AI coding subscription. Delegating a task between Claude Code, Codex, Grok, Kimi, GLM, MiniMax, or a local model still means copy-pasting the prompt, watching another CLI, and pasting the answer back. That friction makes cross-model review easy to skip.

**subagent-model-routing** makes that handoff explicit with three pieces:

Six thin CLI shims (codex-shim.sh, claude-shim.sh, kimi-shim.sh, opencode-shim.sh, grok-shim.sh, and qwen-shim.sh) backed by a shared local Python runtime. They dispatch a prompt to another agentic CLI and return the answer with a SHIM-DONE exit= sentinel, while retaining private run records for inspection and recovery.
A routing skill that defines when to delegate, which model to use, and how to phrase the dispatch, plus tripwire guardrails that catch silent delegation failures (missing sentinel, nonzero exit, truncated output) before they enter your context as usable results.
A model ledger that records every dispatch's wall time, exit code, and outcome, so routing decisions improve from your own observations rather than someone else's defaults.

It's built for **Claude Code, Codex, and GitHub Copilot CLI users** who want to route work through the other installed agentic CLIs. Each host stays native for its own model family: Claude Code uses native Claude age