critique
Skill0 starsChanged 3 months agoMITView on GitHub
Adversarial review of a deliverable against its plan and the original request. Use before declaring any M/L-tier work done, after each execution phase, and whenever the user says review, check this, or is this good — even when the work looks clean, especially when it looks clean.
79 other projects use a version of this.
Adopted by
Most recently taken up by
- Lucface/claude-skillstheir own version, first seen Nov 16, 2026
- Integralist/agent-skillstheir own version, first seen Nov 14, 2026
- theexperiencecompany/gaiatheir own version, first seen Nov 13, 2026
- bas1l/claude-skillstheir own version, first seen Nov 11, 2026
- pche-broken-artist/forge-of-thoughttheir own version, first seen Nov 10, 2026
- BSchrift/critique.nvimtheir own version, first seen Nov 8, 2026
Used with
Projects with it also use delight, teach-impeccable and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
HK-hubSkilldelight
Add moments of joy, personality, and unexpected touches that make interfaces memorable and enjoyable to use. Elevates functional to delightful. Use when the user asks to add polish, personality, animations, micro-interactions, delight, or make an interface feel fun or memorable.
Also in 32 projects that use this
HK-hubSkillteach-impeccable
One-time setup that gathers design context for your project and saves it to your AI config file. Run once to establish persistent design guidelines.
Also in 18 projects that use this
zwy123zwySkillnormalize
将用户随意输入的提示词规范化为结构清晰、便于大模型执行的标准化提示词。适用于需要整理、统一提示词格式时调用。
Also in 18 projects that use this
owshq-mecSkilllayout
Improve layout, spacing, and visual rhythm. Fixes monotonous grids, inconsistent spacing, and weak visual hierarchy. Use when the user mentions layout feeling off, spacing issues, visual hierarchy, crowded UI, alignment problems, or wanting better composition.
Also in 8 projects that use this
- +8 more from HK-hub
Other versions79 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 79
- google-gemini/gemini-cli★ 109ksame name
- yournextstore/yournextstore★ 5.5ksame name
- plugin87/ux-ui-agent-skills★ 3.2ksame name
- remorses/critique★ 1.3ksame name
- educlopez/ui-craft★ 441same name
- nexmoe/DayBee★ 424same name
- nexmoe/eve★ 424same name
- Southclaws/storyden★ 351same name
- 10xChengTu/Mux0★ 337same name
- theexperiencecompany/gaia★ 327same name
69 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository20 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- brainstormSkill
- cleanupSkill
- context-managementSkill
- debuggingSkill
- design-decisionsSkill
- diff-managementSkill
- fan-outSkill
- final-checkSkill
- handoverSkill
- loopsSkill
- memorySkill
- orientSkill
- planningSkill
- researchSkill
- roadmapSkill
- verificationSkill
- web-testingSkill
- writing-reviewSkill
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 72 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Exit criteria: Run every exit criterion from the plan literally, using the verification skill's evidence standards.
- Distinctive: line 32
Never soften findings to protect the work.
“Common” is measured: kept by at least 6 in 10 of the 72 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Critique
The author is the worst judge of their own work. Critique is an attack, not a proofread: actively try to make the work fail.
Who critiques
- Substantial work: spawn a fresh subagent as critic. The role ships in-repo as the
criticagent (.claude/agents/critic.md) — use it (or brief a general agent with its contents) rather than improvising a reviewer prompt. Give it the deliverable, the plan's exit criteria, and the original request — NOT your reasoning or excuses. Fresh eyes only. - Small work: self-critique, but switch stance: assume the work is wrong and hunt for proof.
The attack checklist
- Requirements audit. Re-read the ORIGINAL request word by word. List every explicit and implicit requirement. Mark each: met / partially met / missed / silently reinterpreted. The last category is the most dangerous.
- Exit criteria. Run every exit criterion from the plan literally, using the verification skill's evidence standards. "Probably passes" = fails.
- Falsification attempt. For each claim the work makes, ask: what input, edge case, or reading would break this? Try at least 3 concrete attacks (empty input, extreme values, hostile reader, missing dependency, wrong environment...).
- Consistency. Does the work contradict itself anywhere? Does it contradict the design decisions on record, or memory?
- Completeness of the invisible. What's NOT there: missing error handling, missing citation, unstated assumption, unhandled case, missing file the plan promised?
- Overreach. Did anything get built that wasn't asked for? Scope creep is a defect too.
- Tier audit. Compare the declared task tier (AGENTS.md Task tiers) against what actually happened: an S-tier task that touched more than 2 files or spawned protocol artifacts was mis-tiered — flag it.
Output format
A findings list, each item: severity (blocker / major / minor / nit), location, what's wrong, evidence. No fixes in the critique itself — fixing happens in the loop (see loops skill).
Rules
- A critique that finds nothing is suspect. Look harder once; if still clean, say "clean after second pass" explicitly.
- Critique findings feed the loops skill. Blockers and majors MUST be fixed or explicitly accepted by the user; minors/nits may be batched or waived with a note.
- Never soften findings to protect the work. The work has no feelings.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

