critique
An aggressive critique persona that provides unfiltered, ego-free, and actionable feedback.
Install
mkdir -p .claude/skills/critique-jakessteve && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/15799" && unzip -o skill.zip -d .claude/skills/critique-jakessteve && rm skill.zipInstalls to .claude/skills/critique-jakessteve
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Brutally honest feedback mode. Rips apart ideas, plans, code, designs, and docs with zero sugarcoating. Use when you need the uncomfortable truth, not encouragement.Key capabilities
- →Identify functional correctness issues and edge cases
- →Assess if a solution addresses a real problem
- →Evaluate maintainability and scalability under load
- →Critique craftsmanship in code, design, or writing
- →Determine opportunity cost of current efforts
- →Provide a ranked priority list of fixes
How it works
The skill adopts a CRITIC persona and applies five lenses to analyze the subject, focusing on functionality, necessity, reality survival, craftsmanship, and opportunity cost.
Inputs & outputs
When to use critique
- →Critique a design document
- →Find vulnerabilities in code
- →Identify project planning flaws
- →Stress test an idea
About critique
Adopts a blunt persona to challenge ideas, plans, and code quality. It identifies functional gaps, edge cases, and architectural weaknesses without politeness or hedging.
Brutally honest feedback mode. Rips apart ideas, plans, code, designs, and docs with zero sugarcoating. Use when you need the uncomfortable truth, not encouragement.
When not to use it
- →When encouragement is needed instead of criticism
- →When a design is already affirmed as good
- →When the goal is to improve the design directly
Limitations
- →It criticizes the work, not the person
- →Every criticism must include a concrete fix or direction
- →It acknowledges genuine excellence to maintain credibility
How it compares
This skill provides an unfiltered, structured critique with specific actionable recommendations, unlike a generic review that might offer polite or unquantified feedback.
Compared to similar skills
critique side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| critique (this skill) | 0 | 6mo | No flags | Intermediate |
| maintain | 0 | 3mo | No flags | Advanced |
| claude | 0 | 7mo | Review | Intermediate |
| orchestrator | 0 | 2mo | No flags | Beginner |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
maintain
Alan-Jowett
>
claude
yuiseki
|
orchestrator
fuze210699
orchestrator — an agent skill by fuze210699.
architect-review
sickn33
Master software architect specializing in modern architecture patterns, clean architecture, microservices, event-driven systems, and DDD. Reviews system designs and code changes for architectural integrity, scalability, and maintainability. Use PROACTIVELY for architectural decisions.
solid-principles
SmidigStorm
Enforce SOLID principles (Single Responsibility, Open/Closed, Liskov Substitution, Interface Segregation, Dependency Inversion) in object-oriented design. Use when writing or reviewing classes and modules.
codex
Lucklyric
Invoke Codex CLI for complex coding tasks requiring high reasoning capabilities. This skill should be invoked when users explicitly mention "Codex", request complex implementation challenges, advanced reasoning, or need high-reasoning model assistance. Automatically triggers on codex-related requests and supports session continuation for iterative development.