Skip to content

ilandahan/AID / skills/reflection

reflection

Verified

Automatic quality check on significant outputs via isolated sub-agent. Spawns reflection-agent for objective evaluation without conversation context bias.

criteria/phase-4-development.yaml

View on GitHub
# Phase 4: Development - Reflection Criteria
# Implementation, coding, testing

phase: 4
name: "Development"
description: "Code implementation, unit tests, integration, documentation"

criteria:
  - id: why_alignment
    name: "WHY Alignment"
    weight: 3
    checks:
      - "Code solves the stated problem from requirements?"
      - "Implementation follows approved architecture?"
      - "Not adding unrequested features?"
    scoring:
      excellent: "Code directly addresses requirements"
      acceptable: "Code mostly aligned, minor extras"
      poor: "Code doesn't match requirements or over-engineered"

  - id: phase_compliance
    name: "Phase Compliance"
    weight: 2
    checks:
      - "Following the approved tech spec?"
      - "Not changing architecture without approval?"
      - "Not deploying to production (that's Phase 5)?"
    scoring:
      excellent: "Strictly following approved design"
      acceptable: "Following design with minor adaptations"
      poor: "Significant deviation from approved architecture"

  - id: correctness
    name: "Code Quality"
    weight: 3
    checks:
      - "Code compiles/runs without errors?"
      - "Logic is correct for the requirements?"
      - "Edge cases handled appropriately?"
      - "No obvious bugs?"
    scoring:
      excellent: "Clean, working code with good edge case handling"
      acceptable: "Working code with minor issues"
      poor: "Bugs, syntax errors, or logic flaws"

  - id: security
    name: "Security"
    weight: 3
    checks:
      - "No hardcoded secrets or credentials?"
      - "Input validation present?"
      - "No SQL injection vulnerabilities?"
      - "No XSS vulnerabilities?"
      - "No other OWASP top 10 issues?"
      - "Proper authentication/authorization?"
    scoring:
      excellent: "Secure code following best practices"
      acceptable: "Basic security present"
      poor: "Security vulnerabilities present"

  - id: completeness
    name: "Implementation Completeness"
    weight: 2
    checks:
      - "All acceptance criteria addressable?"
      - "Tests included (if requested)?"
      - "Error handling present?"
      - "WHY documentation in code?"
    scoring:
      excellent: "Complete implementation with tests and docs"
      acceptable: "Core functionality complete"
      poor: "Missing key functionality or tests"

  - id: qa_gate
    name: "QA Gate Validation"
    weight: 3
    description: "Task passed QA sub-agent validation before proceeding"
    checks:
      - "QA criteria file exists for this task (.aid/qa/{task-id}.yaml)?"
      - "All must_achieve criteria satisfied?"
      - "No must_not violations?"
      - "Scope boundaries respected (not_included)?"
      - "Best practices followed?"
    scoring:
      10: "All criteria passed on first review"
      8: "All criteria passed after 1 revision"
      6: "All criteria passed after 2 revisions"
      4: "Required 3 revisions"
      2: "Escalated to human review"
      0: "QA gate bypassed or skipped"
    notes:
      - "QA sub-agent reviews with isolated context (no tech spec)"
      - "Hard block - cannot proceed until PASS"
      - "Max 3 review cycles before escalation"

threshold: 7
max_iterations: 3

notes:
  borderline: "Review edge cases and add security validation"
  improvement_focus:
    - "Add input validation"
    - "Include error handling"
    - "Add WHY comments to complex logic"
    - "Ensure test coverage"
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall