coldeye

Cold-eye

An installable skill for AI agents.

Check whether a newcomer can understand and use what you are about to ship.

Cold-eye is a readiness pass on anything shipped, deemed done, or ready to ship. An agent reads the skill and writes a critique. Verdict first, then a ranked list. The subject does not change.

Get started: pick the agent, install the skill, then run a named example — coldeye.dev/docs/install.

• Cursor
• Claude Code
• Claude.ai

Site: coldeye.dev

First run

Save until-ready.md (or the checklist below), then ask:

Use Cold-eye on until-ready.md. Follow the installed Cold-eye skill. Write the critique. Leave the checklist unchanged.

# Review until ready

Fictional procedure for desk-stamp notes. Labeled example.

1. Open `notes.md`.
2. Read every section.
3. Write findings next to the file as `notes.review.md`.
4. Repeat the review until ready.

The critique lands in until-ready.cold-eye.md. The checklist should be unchanged. A finding should be able to point at the unresolved “until ready” condition. Do not treat one exact failing verdict as proof of loading.

A writable workspace is required. Package coldeye, skill folder cold-eye.

Other installation methods

npm supplies the skill files. It does not register the skill with the agent.

pnpm add -D coldeye

Copy node_modules/coldeye/skills/cold-eye/ into the same destination the Get started page names for your agent.

Updating the npm dependency does not refresh a folder you already copied. Copy again after you bump the package.

License

MIT. Copyright Catalyst Forge LLC.