SkillAgentSearch skills...

audit-workflow

Phase completion verification workflow - triggers brutal audits after phase completion

Install / Use

npx skills add beettlle/pi-spine

Installs into whichever agent you are using.

About this skill
📐

Cursor Rules

Cursor IDE rules (v2)

Quality Score

56/100

Category

Automation

Supported Platforms

Cursor

description: "Phase completion verification workflow - triggers brutal audits after phase completion" globs: [] alwaysApply: false

Phase Completion Audit Workflow

This rule provides guidance for running brutal audits after completing a development phase or milestone.

When to Run Audits

Apply this workflow when:

  • The user mentions completing a "Phase" or "Milestone"
  • The user asks to "verify phase completion"
  • The user wants to ensure code quality before proceeding to the next phase
  • Code quality concerns arise during development

Available Audit Workflows

The repository includes comprehensive brutal audit workflows:

Swift/iOS Projects (Swift 5.9)

  • File: swift-5-9-brutal-audit.mdc
  • Use when: Swift 5.9 language mode (platform minimum is project-defined)
  • Covers: Architectural patterns, concurrency, performance, test quality, Swift 6 readiness, UI compliance

Swift/iOS Projects (Swift 6)

  • File: swift-6-brutal-audit.mdc
  • Use when: Swift 6 language mode (platform minimum is project-defined)
  • Covers: Architectural patterns, strict concurrency compliance, performance, test quality, UI compliance

Python Projects

  • File: python-3-brutal-audit.mdc
  • Use when: Working on Python 3.13+ projects
  • Covers: Async/await patterns, layer boundaries, performance, test quality, type safety, web compliance

Java Projects

  • File: java-brutal-audit.mdc
  • Use when: Working on Java 17 or Java 21 LTS projects
  • Covers: Concurrency (ExecutorService/Virtual Threads), layer boundaries, Effective Java patterns, Optional/Records, test quality, JVM tuning

Go Projects

  • File: go-1-21-brutal-audit.mdc
  • Use when: Working on Go 1.21+ projects
  • Covers: Effective Go, concurrency, testing, structure

JavaScript/TypeScript Projects

  • File: javascript-3-brutal-audit.mdc
  • Use when: Working on JavaScript/TypeScript projects
  • Covers: Modern ES, tooling, testing

Rust Projects

  • File: rust-brutal-audit.mdc
  • Use when: Working on Rust (edition 2021 or 2024) projects
  • Covers: Error handling, ownership, async/concurrency, testing, API guidelines, unsafe

Workflow Steps

  1. Identify the project type (Swift/iOS, Python, Java, Go, JavaScript/TypeScript, or Rust)

  2. Invoke the appropriate audit by mentioning:

    • "run the audit"
    • "check for anti-patterns"
    • "verify phase completion"
    • "run the Swift audit", "run the Python audit", "run the Java audit", "run the Go audit", "run the Rust audit", or "run the JavaScript audit"
  3. Review the audit results — output depends on scope:

    Full phase-completion audit (default when the user says "run the audit" or "verify phase completion" without naming a section):

    • Per-violation report: severity (CRITICAL/HIGH/MEDIUM/LOW), file, lines, recommended fix, effort estimate
    • Section G from the invoked language audit: cleanliness score table (G1), Top 10 files (G2), technical debt tracker (G3)
    • Exit line: Ready for Phase [X+1]? YES/NO

    Section-specific audit (e.g. "Run Section A"):

    • Output only for the requested section(s)
    • Section G deliverables are not required unless Section G is explicitly requested or a full audit is run

    All *-brutal-audit.mdc files share Section G structure; see the invoked audit file for the exact template.

  4. Address blockers — before proceeding to the next phase (applies to full audits, not sectional spot-checks):

    • Fix all CRITICAL and HIGH severity issues
    • Document any acceptable technical debt
    • Ensure the full audit concludes with "Ready for Phase [X+1]? YES"

Integration with Development Standards

These audits complement the always-on rules:

  • Always-on rules (general-llm-anti-patterns.mdc, language-specific standards) prevent issues during development
  • Brutal audits provide deep, milestone-based validation after phase completion

Both work together to ensure code quality throughout the development lifecycle.

Best Practices

  • Run audits after each major phase completion
  • Don't skip audits even if code "looks good"
  • Use audit results to track technical debt over time
  • Address CRITICAL issues before proceeding to next phase
  • Keep audit reports for regression tracking

See Also

  • RULE_COMPOSITION.md - Standard section order and rule structure
  • swift-5-9-brutal-audit.mdc - Swift 5.9/iOS comprehensive audit
  • swift-6-brutal-audit.mdc - Swift 6/iOS comprehensive audit
  • python-3-brutal-audit.mdc - Python comprehensive audit
  • java-brutal-audit.mdc - Java 17/21 comprehensive audit
  • go-1-21-brutal-audit.mdc - Go 1.21+ comprehensive audit
  • javascript-3-brutal-audit.mdc - JavaScript/TypeScript comprehensive audit
  • rust-brutal-audit.mdc - Rust comprehensive audit
  • critical-rules-quick-reference.mdc - Quick reference for when to run audits
  • Language-specific development standards files for ongoing guidance

Related Skills

View on GitHub
GitHub Stars0
CategoryAutomation
UpdatedNaNy ago
Forks0

Security Score

68/100

Audited on Invalid Date

2 medium1 low