Does it detect unused or legacy code?

Yes, it performs a specific Dead Code Audit to find fallback logic, legacy comments, and orphaned tests that should have been removed during refactoring.

What is the Zero Speculation Principle?

It is a core requirement of this skill that every claim (e.g., 'Code implements X') must be accompanied by direct evidence, such as a file path, line number, or specific command output.

How does the skill handle 'meaningless' tests?

The skill includes a mandatory Test Quality Audit that identifies and rejects tautological tests, such as those that only verify compiler-checked facts or mock behavior instead of real logic.

When should I use the review-implementation skill?

You should use it after hyperpowers:executing-plans has completed all tasks, but before you claim the work is finished or move to the finishing-a-development-branch step.

Can I skip certain steps in the review process?

No, the skill operates under a LOW FREEDOM rigidity level, meaning the 4-step review process must be followed exactly to ensure high-quality outcomes.

Implementation Review Agent

Name: Implementation Review Agent
Author: withzombies

bywithzombies

•

Security & Testing

Verifies completed code implementations against specifications using high-scrutiny, evidence-based audits to ensure production readiness.

The review-implementation skill is a high-rigidity auditing tool designed for the Hyperpowers framework. It acts as a final quality gate after a development plan is executed, applying Google Fellow-level scrutiny to ensure implementations match the original 'bd' epic specifications. It enforces a 'Zero Speculation Principle' where every claim must be backed by direct evidence and confidence scores. Key features include a dedicated dead code audit to remove refactoring remnants and a strict test quality gate that rejects tautological or meaningless tests, ensuring that the final output is clean, maintainable, and safe for production.

Key Features

01Automated Dead Code & Refactoring Remnant Audit

02Confidence Scoring system (0.0-1.0) for all findings

03Integrated quality checks for safety, error handling, and anti-patterns

04Evidence-Based Verification with mandatory file/line citations

05Strict Test Quality Gate that rejects 'tautological' tests

0658 GitHub stars

Use Cases

01Self-auditing code for production readiness and anti-patterns

02Finalizing a development branch after all tasks are marked complete

03Validating that new test suites catch real bugs rather than just inflating coverage

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add withzombies/hyperpowers review-implementation

For use in Claude.ai and ChatGPT

Download Skill