taste-to-checks — Claude Code skill: install and use
taste-to-checks is a Claude Code skill that turns repeated review fixes into tests, lint rules, hooks, CI steps, or project rules. It activates after the same miss returns.
MIT ★ 741 updated 2026-11-11
/plugin marketplace add MiniMax-AI/mcode-stack
/plugin install ronny@mcode-stackWhat taste-to-checks does
taste-to-checks turns a repeated code-review correction into an executable rule. It helps phrase the rule as one testable sentence, then places it in a test, lint rule, hook, CI step, AGENTS.md, or another skill according to the rule's shape.
The check is run against the old mistake first, where it should fail with a clear message, and then against current main, where it should pass. Code patterns can use a linter or CI grep; required behavior belongs in a test; actions required on every commit belong in a hook. Text in AGENTS.md can be ignored, while a failing check records the violation.
Its output contains the rule sentence, its destination, the command that failed on the old version and passed on main, plus the PR link. The workflow keeps one rule per PR.
What it helps with
- turn a repeated review comment into an executable check
- block new
anyusage inpackages/core - require a test that fails on the buggy version
- choose between a test, lint rule, hook, and
AGENTS.md - verify a check against the old mistake and current
main
Example prompts for the agent
- Turn the repeated import-order comment into a testable rule and choose where it should run.
- Define a check for new
anyinpackages/core, then run it against the old mistake and currentmain. - A second PR missed the same test requirement. Show the rule, its destination, and the command that proves it catches the old diff.
How to install taste-to-checks
1Plugin marketplace
/plugin marketplace add MiniMax-AI/mcode-stack/plugin install ronny@mcode-stackThis installs the whole ronny plugin with all its skills
2Skills CLI (Vercel)
npx skills add MiniMax-AI/mcode-stack --skill taste-to-checks -a claude-code3Manually: put the files in place
~/.claude/skills/taste-to-checks/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/taste-to-checks/
1Skills CLI (Vercel)
npx skills add MiniMax-AI/mcode-stack --skill taste-to-checks -a codex2Manually: put the files in place
.agents/skills/taste-to-checks/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add MiniMax-AI/mcode-stack --skill taste-to-checks -a cursor2Manually: put the files in place
.cursor/skills/taste-to-checks/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install MiniMax-AI/mcode-stack stacks/ronny/skills/taste-to-checks2Manually: put the files in place
.github/skills/taste-to-checks/SKILL.md1Skills CLI (Vercel)
npx skills add MiniMax-AI/mcode-stack --skill taste-to-checks -a gemini-cli2Platform CLI
gemini skills install https://github.com/MiniMax-AI/mcode-stack --path stacks/ronny/skills/taste-to-checksManually: put the files in place
.devin/skills/taste-to-checks/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about taste-to-checks
How do I install taste-to-checks in Claude Code?
Add the marketplace with /plugin marketplace add MiniMax-AI/mcode-stack, then install the skill with /plugin install ronny@mcode-stack.
What does taste-to-checks do?
It converts a review correction that keeps recurring into a test, lint rule, hook, CI step, or project rule. It then lays out a check against the old mistake and current main.
Where should a rule from taste-to-checks go?
Use a lint rule or CI grep for forbidden code patterns, a test for required behavior, and a hook for actions on every commit. Context-dependent guidance can live in AGENTS.md or a skill.
Source text
SKILL.md · MIT
--- name: taste-to-checks description: Turn a review correction we keep making on agent output into something that runs on its own, a test, lint rule, hook, CI step or project rule, so the miss does not come back next session. Use when we keep correcting agents on the same mistake (import order, naming, a forbidden API, a missing test), when the same review comment shows up on a second PR, or after a review finds a miss that a check could have caught. --- # Taste to Checks If we tell an agent the same thing twice, the second time is our fault. Taste that lives only in our heads has to be applied by hand in every review, and it does not scale with the number of sessions we run. We write it down where it runs by itself. ## 1. Catch the repeat Signs a correction should become a check: - we wrote the same review comment on two agent PRs; - the same thing went wrong after compaction or in a new session; - `evidence-first-review` found a miss that a mechanical check would catch. One-off quirks stay out. Only encode what has happened more than once or caused real damage. ## 2. Write the rule as one testable sentence Bad: "keep the code clean". Good: "no new `any` in `packages/core`", "every new tool has a description under 200 characters", "a bug fix PR includes a test that fails on main". If we cannot make it testable, it is a judgment call. Keep it in review. ## 3. Pick the strongest place it fits | Rule shape | Where it goes | | --- | --- | | Code pattern to forbid | lint rule or a grep in CI, with an error message that names the fix | | Behaviour that must hold | test, written to fail on the bad version first | | Must happen on every edit or commit | hook (pre-commit, or a MiniMax Code hook) | | Needs context, not mechanical | project rules (`AGENTS.md`) or a skill, as a last resort | Prefer the top of the table. Text in `AGENTS.md` is read and sometimes ignored; a failing check is not. ## 4. Prove the check works Run it against the past mistake: check out the commit or reapply the bad diff and confirm it fails with a clear message. Then confirm it passes on current main. A check that never failed has not been shown to work. ## 5. Ship it small One rule per PR, with the review comment or incident it came from linked in the description. ## Output The rule sentence, where it went, the command that showed it failing on the old mistake and passing on main, and the PR link. ## References - Addy Osmani, ["Critical thinking when agents do the work"](https://x.com/addyosmani/status/2117512471930167377)
© Ronny (MiniMax Code team) · MIT · MiniMax-AI/mcode-stack