wernerhe/super-model/super-code-review

Trust tier 2
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Run a comprehensive code review at a milestone (typically before merging a feature branch). Writes an HMAC-signed cache marker for super-prepare-branch to consume; if nothing changed since the last review, super-prepare-branch can skip re-running.

Install

npx vskill@latest install wernerhe/super-model/super-code-review

SKILL.md

super-code-review

The comprehensive code review skill. Catches the issues that fast feedback loops miss - dead code, namespace drift, dependency rot, coverage gaps - by sweeping the diff systematically across 9 dimensions.

When to use

  • A feature branch is "done" and the user is about to push / open PR / merge.
  • A risky refactor needs a once-over before going downstream.
  • A super-prepare-branch invocation finds no cache marker and chain-fires this skill.

The 9-module sweep

  1. file-references - dead imports, orphan files, broken doc links.
  2. namespace-consistency - new identifiers fit existing naming conventions.
  3. dead-code - functions / variables / imports defined but never used in the diff.
  4. test-coverage - changed lines covered by tests; new branches have tests.
  5. security-scan - secrets in plaintext, command injection, dependency vulns.
  6. dependency-health - new deps vetted, version pins reasonable, no abandoned packages.
  7. doc-drift - public API changes reflected in docs / READMEs / docstrings.
  8. performance-hotspots - O(n^2) inside loops, accidental N+1, repeated regex compiles.
  9. complexity - functions over a threshold (cyclomatic / cognitive), single-responsibility violations.

Each module produces zero or more findings tagged with severity (critical / important / minor) and an optional fix proposal.

Cache marker output

On successful completion, writes:

<project>/.super/cache/code-review-<commit-sha>.json

via super_lib.cache.write_marker. The payload includes:

  • verdict: READY-TO-MERGE | NEEDS-FIXES.
  • findings: list of findings with severity, file, line range, description, suggested fix.
  • commit_sha: HEAD SHA at review time.
  • timestamp: UTC ISO-8601.

HMAC-signed so a downstream super-prepare-branch can trust the verdict; tampered markers are treated as missing and trigger a re-review.

Hard rules

  • NEVER claim "READY-TO-MERGE" with unresolved CRITICAL findings.
  • ALWAYS write the cache marker, even on NEEDS-FIXES verdict - super-prepare-branch needs the record.
  • NEVER auto-apply fixes from modules marked fix_policy: never. Always present and ask.

Verify after running

  • <project>/.super/cache/code-review-<sha>.json exists with a valid HMAC.
  • The verdict matches the user's assessment (no over-stating "ready").
  • Each finding has a clear file:line and a description the user can act on.

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/wernerhe/super-model/super-code-review/badge)](https://verified-skill.com/skills/wernerhe/super-model/super-code-review)

Something wrong with this skill or its page? Tell us.