skeptical_review

Perform a critical, skeptical code review of the project from the perspective of a Principal Engineer.

12 stars

bylionbenjamin

View on GitHub Installation ↓

Best use case

skeptical_review is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Perform a critical, skeptical code review of the project from the perspective of a Principal Engineer.

Teams using skeptical_review should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

You only need a quick one-off answer and do not need a reusable workflow.
You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/skeptical_review/SKILL.md --create-dirs "https://raw.githubusercontent.com/lionbenjamin/agent-templates/main/skills/skeptical_review/SKILL.md"

Manual Installation

Download SKILL.md from GitHub
Place it in .claude/skills/skeptical_review/SKILL.md inside your project
Restart your AI agent — it will auto-discover the skill

How skeptical_review Compares

Feature / Agent	skeptical_review	Standard Approach
Platform Support	Not specified	Limited / Varies
Context Awareness	High	Baseline
Installation Complexity	Unknown	N/A

Frequently Asked Questions

What does this skill do?

Perform a critical, skeptical code review of the project from the perspective of a Principal Engineer.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

Related Guides

AI Agents for Coding

Browse AI agent skills for coding, debugging, testing, refactoring, code review, and developer workflows across Claude, Cursor, and Codex.

Best AI Skills for Claude

Explore the best AI skills for Claude and Claude Code across coding, research, workflow automation, documentation, and agent operations.

SKILL.md Source

# Skeptical Review Skill

Critical code review from an experienced Principal Engineer.

## When to Activate

This skill is relevant when:
- Reviewing code for long-term maintainability
- Assessing architectural decisions
- Evaluating AI-generated or rapidly-produced code
- Checking for over-engineering
- Pre-production readiness review

## Core Principles

### Critical Lens
- Experienced with 20+ years perspective
- Focused on long-term maintainability
- Question decisions, not just validate
- High standards for production code

### Anti-Magic
- Dislike unnecessary abstractions
- Avoid over-engineering
- Simple beats clever
- Code should be boring and obvious

### Maintainability Focus
- Will someone understand this in 6 months?
- Is debugging straightforward?
- Can this be tested reliably?
- Does complexity match problem scope?

### Production Readiness
- What breaks first under load?
- How will this fail in edge cases?
- Is error handling robust?
- Can we operate this?

## Quick Checks

When performing skeptical review, verify:
- [ ] Architecture: Coherent design exists
- [ ] Architecture: Clear separation of concerns
- [ ] Architecture: Appropriate patterns for problem size
- [ ] Code Quality: Naming is clear and consistent
- [ ] Code Quality: Error handling is comprehensive
- [ ] Code Quality: No code smells
- [ ] Agentic Issues: No unnecessary abstractions
- [ ] Agentic Issues: Patterns are consistent throughout
- [ ] Agentic Issues: No forgotten TODO comments
- [ ] Agentic Issues: No copy-paste artifacts
- [ ] Agentic Issues: Solutions match problem complexity
- [ ] Dependencies: All dependencies justified
- [ ] Dependencies: No bloat or "just in case" libraries
- [ ] Dependencies: Complexity proportional to value
- [ ] Production Risks: Load handling considered
- [ ] Production Risks: Edge cases handled
- [ ] Production Risks: Failure modes identified
- [ ] Production Risks: Monitoring and debugging possible
- [ ] Testability: Code is unit-testable
- [ ] Testability: Integration points are clear
- [ ] Testability: Mocking/stubbing is feasible

## Review Focus Areas

### Architecture & Structure
- Is there a coherent design?
- Does it feel maintained or just assembled?
- Are responsibilities clear?
- Is coupling appropriate?

### Code Quality
- Naming clarity and consistency
- Pattern usage and appropriateness
- Error handling completeness
- Separation of concerns

### Signs of Agentic Sloppiness
- Unnecessary abstractions
- Inconsistent patterns across files
- Abandoned TODO comments
- Copy-paste artifacts
- Over-complicated solutions to simple problems
- Generic variable names
- Missing error handling

### Dependencies & Complexity
- Are dependencies justified?
- Is complexity proportional to problem?
- Could this be simpler?
- Is there vendor lock-in?

### Production Failure Prediction
- What breaks first under load?
- Where are the edge cases?
- What happens when external services fail?
- Are race conditions possible?

## Output Format

Provide prioritized improvements:
- **Highest-Impact First**: Focus on what matters most
- **Concrete Suggestions**: Show what better looks like
- **Specific Examples**: Include code refactors or pattern changes
- **Not Vague Advice**: Actionable, not philosophical
- **Balanced Perspective**: Acknowledge good parts too