Engineering standards and code governance enforcement with parallel agent execution. Triggers during code reviews, commits, and standards discussions. 70-90% token savings via delegated execution.

2 stars
0 forks
JavaScript
16 views

SKILL.md


name: governance version: "2.0.0" description: "Engineering standards and code governance enforcement with parallel agent execution. Triggers during code reviews, commits, and standards discussions. 70-90% token savings via delegated execution." category: development tags:

  • governance
  • code-review
  • standards
  • compliance
  • quality
  • automation author: wrsmith108 repository: https://github.com/wrsmith108/governance-claude-skill license: MIT triggers: keywords:
    • code review
    • review this
    • commit
    • before I merge
    • standards
    • compliance
    • code quality
    • best practices
    • retro
    • retrospective explicit:
    • /governance
    • /review
    • /retro
    • /audit delegates_to: governance-specialist allowed-tools:
  • Bash
  • Read
  • Write
  • Edit
  • Grep
  • Glob

Governance Skill

Enforces engineering standards and code quality policies with parallel agent execution for context efficiency.

Standards Document: docs/architecture/standards.md Audit Script: npm run audit:standards


Execution Model (v2.0)

Type: Delegated Execution

This skill operates through the governance-specialist subagent for context efficiency:

  1. Spawn the specialist agent in parallel
  2. Delegate the governance task with clear instructions
  3. Await completion without intervention
  4. Report the structured summary to the user

Token Efficiency

Execution Mode Token Usage Benefit
Direct (v1.x) ~3,000-8,000 Full context in main thread
Delegated (v2.0) ~500-800 70-90% token savings

How to Invoke (Orchestrator Instructions)

When this skill is triggered, delegate to a parallel agent:

Code Review

Task({
  description: "Governance code review",
  prompt: `Review the following files for standards compliance:
- [list files or PR description]

Execute full code review workflow:
1. Run standards audit
2. Check each file against standards
3. Fix ALL issues immediately (zero deferral)
4. Create code review report
5. Return structured summary`,
  subagent_type: "reviewer"
})

Pre-Commit Check

Task({
  description: "Pre-commit governance",
  prompt: `Run pre-commit governance checks:
1. typecheck, lint, format:check, test
2. Check for untracked source files
3. Return pass/fail with specific failures`,
  subagent_type: "reviewer"
})

Retrospective

Task({
  description: "Run retrospective",
  prompt: `Run retrospective for [topic/sprint]:
1. Analyze completed issues
2. Gather metrics
3. Write report
4. Return summary`,
  subagent_type: "reviewer"
})

Behavioral Classification

Zero Deferral Policy: The specialist fixes ALL issues immediately - no tickets, no "later".

Pattern Status
"Would you like me to fix these issues?" Anti-pattern
"Created ticket to track this for later." Anti-pattern
"Found 5 issues. All fixed. Commits: abc123, def456." Correct

Quick Start

1. Run Setup Check

node .claude/skills/governance/scripts/governance-check.mjs

2. Two-Document Model

Document Purpose Update Frequency
CLAUDE.md AI operational context Frequent
standards.md Engineering policy (authoritative) Infrequent

3. Common Operations

# Check standards compliance
npm run audit:standards

# Pre-commit suite
npm run typecheck && npm run lint && npm run test

Core Standards Quick Reference

Code Quality (§1)

Standard Requirement Check
TypeScript Strict mode enabled npm run typecheck
No any types Use unknown for external data Audit script
Max file length 500 lines Audit script
JSDoc Required for public APIs Audit script

Naming Conventions (§1.2)

Element Convention Example
Components PascalCase Button.tsx
Utilities camelCase userService.ts
Constants SCREAMING_SNAKE MAX_RETRIES
DB columns snake_case user_id

Testing (§2)

Layer Target
Unit tests 90-100%
API routes 90-100%
E2E critical 100%

Compliance Check

npm run audit:standards

Checks:

  • TypeScript strict mode
  • No any types in source
  • File length limits (500 lines)
  • JSDoc on exported functions
  • Test files exist

Anti-Patterns vs Correct Patterns

Code Quality

Anti-Pattern Correct Pattern
any type unknown for external data
❌ 600+ line files ✅ Split at 500 lines
❌ Hardcoded secrets ✅ Environment variables

Workflow

Anti-Pattern Correct Pattern
❌ Commit to main ✅ Feature branches with PR
❌ Skip audit ✅ Run audit:standards before commit
❌ Defer issues ✅ Fix immediately

Pre-Commit Checklist

[ ] npm run typecheck
[ ] npm run lint
[ ] npm run test
[ ] npm run audit:standards
[ ] No console.log statements
[ ] No hardcoded secrets

PR Review Checklist

Category Criteria
Correctness Does it work? Edge cases?
Security No secrets, injection, auth bypasses
Performance No N+1, memory leaks
Testing Tests exist for new code

Output Format (Specialist Response)

The governance-specialist returns structured summaries:

- **Task:** Code review for feature X
- **Checks Run:** typecheck, lint, audit:standards
- **Results:**
  - Found 3 issues: 1 missing test, 1 type issue, 1 style
  - All issues fixed in-place
- **Issues Fixed:** 3 (commits: abc1234, def5678, ghi9012)
- **Report:** docs/code_review/2026-01-28-feature-x.md
- **Status:** PASS

Updated: January 2026 Version 2.0.0 - Delegated Execution Model