code-review
Code review practices with technical rigor and verification gates. Use for receiving feedback, requesting code-reviewer subagent reviews, or preventing false completion claims in pull requests.
85
83%
Does it follow best practices?
Impact
82%
1.17xAverage score across 3 eval scenarios
Passed
No known issues
Evaluation results
90%
7%Respond to External Code Review
Code review reception
No performative agreement
100%
100%
No gratitude expressions
100%
100%
Stops for unclear item
100%
100%
No partial implementation plan
0%
0%
YAGNI check on repository interface
53%
100%
Pushback on Comment 3
100%
100%
Technical reasoning in pushback
100%
100%
Accepts valid security comments
100%
100%
No implementation before verification
100%
100%
Blocking issues prioritized
100%
100%
87%
Fix and Verify a Cart Calculation Bug
Verification before completion claims
Ran bun test
0%
0%
Actual command shown
100%
100%
Actual output included
100%
100%
Test count stated
100%
100%
No premature success claim
100%
100%
No 'should/probably/seems' language
100%
100%
Bug fix is correct
100%
100%
Requirements checklist
70%
70%
No unsupported claims
100%
100%
70%
30%Set Up Code Review After Feature Implementation
Code reviewer subagent dispatch
BASE_SHA via git
0%
0%
HEAD_SHA via git
0%
0%
Actual SHA values
100%
100%
WHAT_WAS_IMPLEMENTED field
70%
100%
PLAN_OR_REQUIREMENTS field
70%
100%
DESCRIPTION field
80%
100%
Critical issues: fix immediately
30%
100%
Important issues: fix before proceeding
0%
100%
Minor issues: noted for later
50%
100%
Task tool dispatch type
0%
0%
- Repository
- secondsky/claude-skills
- Commit
90d6bd7
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.