agentleFS
Sign inStart free
LucasCooperBeyLucasCooperBey/claude-skills/skills/critique/SKILL.md

critique

Skill1 starsChanged 5 months agoMITView on GitHub

Single-pass 3-agent critique for ANY work product — plans, code, designs, architecture. The general-purpose refinement tool.

We split this skill into the rules it gives an agent

5 rules. 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 72 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If --type not specified, auto-detect:

    line 37
  2. Distinctive:

    Use the Agent tool (subagent_type: general-purpose) to spawn 3 critique agents.

    line 78
  3. Distinctive:

    If yes, revise the work product inline

    line 115
  4. Distinctive:

    After writing a plan (before executing)

    line 119
  5. Distinctive:

    After drafting architecture or design docs

    line 120

“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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Critique (/critique)

Single-pass 3-agent critique that works on any work product — plans, code, designs, architecture, specs, strategies.

Unlike /simplify (code-only), this works on everything.

Arguments

ArgEffect
--type TYPEForce type: code, plan, design, architecture, strategy (auto-detected if omitted)
--agents "A,B,C"Custom agent perspectives
--securityPreset: Solution-Fit, Security, Edge Cases
--performancePreset: Solution-Fit, Performance, Elegance
--applyAuto-apply improvements (default: present findings and ask)

Instructions

Step 1: Identify the Work Product

Find what to critique from context:

  • Plan file in conversation or referenced path
  • Code in recent conversation or uncommitted changes
  • Design/architecture document
  • Any substantive work product

If nothing found:

"Nothing to critique. Point me at a file, paste content, or describe what to review." Then STOP.

Step 2: Detect Type

If --type not specified, auto-detect:

  • Has code blocks with implementation → code
  • Markdown with ## headers, task lists, steps → plan
  • Discusses components, interfaces, data flow → architecture
  • UI/UX layouts, wireframes, component specs → design
  • Business goals, roadmap, priorities → strategy
  • Default: plan

Key difference from /simplify: default is plan, not code.

Step 3: Select Agent Perspectives by Type

For plans:

  1. Completeness — Are all steps defined? Missing prerequisites? Gaps between steps?
  2. Feasibility — Can each step actually be executed? Unrealistic assumptions?
  3. Risk & Edge Cases — What could go wrong? What's the fallback?

For code:

  1. Solution-Problem Fit — Does it solve the stated problem?
  2. Edge Cases & Gaps — What could break?
  3. Elegance & Simplicity — Can it be simpler?

For architecture:

  1. Separation of Concerns — Clean boundaries? Right abstractions?
  2. Scalability & Evolution — Will this grow well? Paint into corners?
  3. Integration Risk — Dependencies, coupling, failure modes?

For design:

  1. User Flow — Is the happy path clear? Dead ends?
  2. Consistency — Patterns reused? Naming coherent?
  3. Edge States — Empty, error, loading, overflow handled?

For strategy:

  1. Goal Alignment — Do actions connect to stated goals?
  2. Prioritization — Is sequencing justified? Dependencies right?
  3. Blind Spots — What's not being considered?

Custom agents override these defaults.

Step 4: Dispatch 3 Agents in Parallel

Use the Agent tool (subagent_type: general-purpose) to spawn 3 critique agents.

Each agent prompt includes:

  • The work product (full text)
  • Their specific critique focus
  • The detected type
  • Project context (CLAUDE.md excerpt if available)
  • Instruction: "Return a numbered list of specific, actionable findings. No fluff. Each finding must include: what's wrong, why it matters, and a concrete fix."

Step 5: Synthesize & Present

Collect all agent findings. Deduplicate. Group by severity:

## Critique Results (TYPE)

### Critical (must fix)
1. [Finding] — [Fix]

### Important (should fix)
1. [Finding] — [Fix]

### Minor (nice to have)
1. [Finding] — [Fix]

### Verdict
X critical, Y important, Z minor findings across 3 perspectives.

Step 6: Apply or Report

If --apply flag or type is code:

  • Apply improvements directly
  • Show diff summary

If type is plan, design, architecture, strategy:

  • Present findings and ask: "Apply these improvements?"
  • If yes, revise the work product inline

When to Use

  • After writing a plan (before executing)
  • After drafting architecture or design docs
  • Quick single-pass code review (lighter than /pp2p)
  • Strategy documents and roadmaps
  • Any time you want 3 perspectives on non-code work

Comparison

SkillPassesWorks OnPolish Phase
/simplify1Code onlyNo
/critique1EverythingNo
/pp2p2EverythingYes
/pp3p3EverythingYes

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free