Open-source MIT skill

Suggest a lighter model tier. Confirm before it runs.

Auto Model Router is a portable policy for Cursor, Claude, Codex, and other multi-model agents. It classifies work into the lightest sufficient tier, shows the suggestion, and waits for you to confirm or override. The skill stays usable without an email—this page is only for optional feedback.

Live demo

Motion replay of the real pass

Play the live-motion chat replay. A Codex-style UI animates the real Auto Model Router pass: load skill → suggest fast → wait → rename getUserName to fetchUserName → suggest fast → wait → confirm → run on fast. This is the flow happening, not a slide explainer. It shows suggest → wait → confirm, not billing or token savings.

Motion replay of the real suggest → wait → confirm pass (Codex-style chat UI). Exact replies are quoted below. This is not a slide explainer or a live IDE embed.

Exact replies from the pass

  1. 1 · Activation Auto suggests fast — no substantial work has been requested yet. Confirm to run, or override: fast / standard / reasoning / max.
  2. 2 · Rename test Auto suggests fast — this is a small, reversible, single-file rename with same-file call sites. Confirm to run, or override: fast / standard / reasoning / max.

Honesty: this demonstrates suggest → wait behavior. It is not a billing dashboard and does not claim usage or cost savings.

Try this in Codex

Paste after the skill is installed. Codex should suggest fast and stop. Do not treat a confirm as implied.

Following the auto-model-router skill in this session.

TEST — do not write code yet.

Task to classify: "Rename the function getUserName to fetchUserName in one file and update its call sites in that same file."

1. Suggest the lightest sufficient tier (fast / standard / reasoning / max) with a one-line reason.
2. STOP and wait for my confirm or override.
3. Do not edit files until I say confirm (or a different tier).

Reply ONLY in this shape first:
Auto suggests <tier> — <reason>. Confirm to run, or override: fast / standard / reasoning / max.

Example flow (illustrative)

How the skill chooses a tier

Illustrative CSS animation of classify → suggest → confirm → run. Separate from the live-motion chat replay above. This mock is not a live IDE recording or Cursor embed.

First example: rename getUserName to fetchUserName. Auto suggests fast because it is a clear rename with low judgment. Confirm, then it runs on the fast tier. Second example: auth refresh fails only for users who signed up before token rotation. Auto suggests reasoning. Confirm, then it runs on the reasoning tier.

  1. Rename getUserName → fetchUserName across the auth module. Auto suggests fast — clear rename, low judgment. Confirm, then it runs on the fast tier.
  2. Auth refresh fails only for pre-rotation users. Auto suggests reasoning — unknown-root-cause debugging. Confirm, then it runs on the reasoning tier.

Get it

Install & listings

Install from GitHub today. Community and official catalogs are listed only when a real public URL exists. The plugin does not guarantee billing or token savings.

Works today

GitHub

Clone the repo or add it as a plugin marketplace. Full host steps stay in INSTALL.md.

cursor.directory · community

cursor.directory

Community listing. It may still be in security scan or unpublished until review finishes. This is not Cursor's official Marketplace.

No public URL yet

Cursor official Marketplace

Submitted — pending Cursor review

No public product page is available yet. This is not a live official Marketplace listing.

GitHub install commands

These add the GitHub repo as a plugin source. They are not official Cursor Marketplace or Anthropic catalog listings.

# Claude Code (in a session)
/plugin marketplace add 1ststepai/auto-model-router
/plugin install auto-model-router@auto-model-router
/reload-plugins
# Codex
codex plugin marketplace add 1ststepai/auto-model-router
codex plugin add auto-model-router@auto-model-router

Cursor Teams / Enterprise: Dashboard → Plugins & MCPs → Import from Repo → https://github.com/1ststepai/auto-model-router, then install Auto Model Router from Customize. Individuals can copy the skill or use the apply script in INSTALL.md.

Classifyfast, standard, reasoning, or max — capability labels, not vendor model names.
Suggest → confirmThe agent explains the tier first. You confirm or override before substantial work.
Escalate onceIf a light attempt fails, it says so and steps up. Explicit model choices always win.

Honest scope

This is a readable heuristic, not trained routing ML and not a billing dashboard. Usage savings happen only if the confirmed lighter tier is mapped to a cheaper or faster model (or lower effort) and that mapped choice is what actually runs. Percentages in the repo estimator are relative comparisons from a local log, not vendor invoices or promised savings.

Leave feedback

Email is optional. The skill works the same either way. If you share an email, you must opt in to hear about this tool and related 1stStep tools.

Leave feedback, an email, or both. At least one is required.

No silent email harvesting. Required only if you want a reply or updates. See the privacy policy.