Every model gets a written job: driver, workhorse, gatekeeper
Give every model a written job in AGENTS.md: driver, workhorse, gatekeeper. Marketer running a Claude Code skills stack.
what to steal
Write each model's role into AGENTS.md so routing isn't in your head.
Our takeaway from the source, not a quote.
ryandoser-agents-md-model-roles.toml
# Give every model a written job in AGENTS.md: driver, workhorse, gatekeeper # source: https://ryandoser.com/is-chatgpt-pro-worth-it/ [rig] who = "Marketer running a Claude Code skills stack" author = "Ryan Doser" date = "2026-10" harness = ["Claude Code", "Codex"] models = ["Claude Opus 5.5 (daily driver)", "Claude Sonnet 5.5 (workhorse)", "GPT-6.1 Sol, high reasoning (gatekeeper via Codex CLI)"] [keep] - Models section in AGENTS.md - Markdown file of Opus 5.5's best outputs that Sonnet reads first - Claude Code Skills Stack
| thing | status | why (per source) |
|---|---|---|
| Gatekeeper role | moved to GPT-6.1 Sol on 2026-10-01 | A critic from another model family catches what a Claude model produces |
Source: https://ryandoser.com/is-chatgpt-pro-worth-it/ · checked 2026-10-14
// similar rigs
~/rigs/mikeholp-route-claude-codex.toml2026-10-11
Route by task type and make each agent review the other
steal → Have the agent that didn't write the change review it.
Claude CodeCodexCursor Claude Opus 5.5Claude Sonnet 5.5 git push / commit in C
by Mike Holp
~/rigs/merlin-claude-manager-codex-….toml𝕏 · 2026-10-14
Claude manages, Codex grinds
steal → Use one model to plan architecture and another to churn through build steps.
Claude CodeCodex Claude Fable as manageGPT-6 Astra on Codex
by @Merlin412188240
~/rigs/stefancrncic-t3-three-machin….toml𝕏 · 2026-10-13
Five harnesses, three machines, one orchestrator
steal → Install several harnesses on worker machines and drive them from one place.
Claude CodeCodexCursor
by Stefan (@stefancrncic)