AgentSkillsCN

review

运行质量门禁，审查当前工作进展

SKILL.md
--- frontmatter
name: review
description: Run quality gates and review current work
allowed-tools: Read, Grep, Glob, Bash

Run quality gates and review current work.

Instructions

  1. •Read .claude/rules/agent-guide.md for project-specific quality gate commands and the Production Code Quality Checklist.
  2. •Read docs/spec/.llm/PROGRESS.md for known patterns and conventions.
  3. •Run quality gates from the agent guide rules (build, test, lint, type-check as applicable).
  4. •Check conventions:
    • •Code follows rules in .claude/rules/ (go-patterns, typescript-patterns, etc.)
    • •New files have appropriate structure and documentation
    • •No hardcoded credentials, secrets, or environment-specific values
    • •Error handling follows project patterns
    • •Implementation matches technical spec (if one exists): data models, API contracts, error codes
  5. •Spec compliance (if a technical spec exists for the changed feature):
    • •Every data model field matches the spec (types, constraints, nullability)
    • •Every API endpoint matches the spec (path, method, request/response shape, status codes)
    • •Every error code in the spec is handled in the implementation
    • •Edge cases documented in the spec have corresponding code paths and tests
  6. •Code quality audit:
    • •All error paths are tested, not just happy paths
    • •Input validation exists at system boundaries (handlers, CLI, config parsing)
    • •No swallowed errors (_ = err without justification)
    • •Functions are < 60 lines; complex logic is commented with WHY
    • •New public packages have doc.go with usage examples
    • •README updated if new user-facing functionality was added
  7. •Security check:
    • •No secrets, tokens, or passwords in code or config files
    • •User input is validated and sanitized before use
    • •SQL queries use parameterized statements (no string concatenation)
    • •File paths from user input are sanitized
  8. •Report results in a structured format:
    • •Quality gates: pass/fail with details
    • •Spec compliance: field-by-field match (or N/A if no spec)
    • •Code quality: issues found with file:line references
    • •Security: issues found or clean
    • •Convention issues: list with file and line
    • •Recommendations: what to fix and how (prioritized)

$ARGUMENTS