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-3-impl-plan.yaml

View on GitHub
# Phase 3: Implementation Plan - Reflection Criteria
# Now includes sub-phase validation (3a, 3b, 3c)

phase: 3
name: "Implementation Plan"
description: "Consolidation, task breakdown, sprint planning, Jira population"

sub_phases:
  - id: "3a"
    name: "Consolidation & Validation"
    description: "Resolve contradictions, create consolidated master document"
  - id: "3b"
    name: "Task Breakdown"
    description: "Break down into actionable tasks with estimates"
  - id: "3c"
    name: "Jira Population"
    description: "Create Jira issues with full information"

criteria:
  # ===== PHASE 3a CRITERIA =====
  - id: consolidation_completeness
    name: "Consolidation Completeness"
    weight: 3
    sub_phase: "3a"
    checks:
      - "All source documents analyzed (PRD, Tech Spec, Research)?"
      - "All contradictions between documents identified?"
      - "No information lost or removed from source documents?"
      - "Consolidated document is self-contained?"
    scoring:
      excellent: "Complete consolidation with all information preserved"
      acceptable: "Minor information gaps, easily fixable"
      poor: "Missing major sections or information lost"

  - id: contradiction_resolution
    name: "Contradiction Resolution"
    weight: 3
    sub_phase: "3a"
    checks:
      - "All contradictions explicitly documented?"
      - "Resolution follows hierarchy (Research > PRD > Tech Spec)?"
      - "Rationale provided for each resolution?"
      - "Original conflicting text preserved for reference?"
    scoring:
      excellent: "All contradictions resolved with clear rationale"
      acceptable: "Most resolved, minor gaps in rationale"
      poor: "Contradictions unresolved or resolutions unjustified"

  - id: consolidation_approval
    name: "User Approval"
    weight: 2
    sub_phase: "3a"
    checks:
      - "Consolidated document presented to user?"
      - "User explicitly approved the consolidation?"
      - "Any user feedback incorporated?"
    scoring:
      excellent: "Explicit user approval received"
      acceptable: "Implicit approval or minor changes pending"
      poor: "No approval or user rejected"

  # ===== PHASE 3b CRITERIA =====
  - id: why_alignment
    name: "WHY Alignment"
    weight: 3
    sub_phase: "3b"
    checks:
      - "Each task traces back to consolidated spec section?"
      - "Task breakdown serves the defined architecture?"
      - "Priority reflects business value from PRD?"
      - "No orphan tasks without spec reference?"
    scoring:
      excellent: "Clear traceability from every task to requirements"
      acceptable: "Most tasks link to specs, minor gaps"
      poor: "Tasks without clear connection to specs"

  - id: phase_compliance
    name: "Phase Compliance"
    weight: 2
    sub_phase: "3b"
    checks:
      - "Planning tasks, not executing them?"
      - "Not writing production code yet?"
      - "Following approved architecture from consolidated spec?"
    scoring:
      excellent: "Pure planning and task definition"
      acceptable: "Mostly planning with minor scaffolding"
      poor: "Contains implementation code"

  - id: task_quality
    name: "Task Quality"
    weight: 3
    sub_phase: "3b"
    checks:
      - "All tasks < 4 hours (prefer 1-2 hours)?"
      - "Each task has clear, testable acceptance criteria?"
      - "Dependencies correctly identified?"
      - "Critical path documented?"
      - "Estimates are realistic based on complexity?"
    scoring:
      excellent: "Well-structured tasks with realistic estimates"
      acceptable: "Good structure with minor sizing issues"
      poor: "Unrealistic estimates or missing dependencies"

  - id: sprint_planning
    name: "Sprint/Agile Planning"
    weight: 2
    sub_phase: "3b"
    checks:
      - "Sprint goals clearly defined?"
      - "Capacity considered in sprint allocation?"
      - "Stories prioritized by business value?"
      - "Dependencies respected in sprint order?"
    scoring:
      excellent: "Realistic sprint plan with clear goals"
      acceptable: "Sprint plan exists with minor capacity issues"
      poor: "No sprint plan or unrealistic allocation"

  - id: risk_assessment
    name: "Risk Assessment"
    weight: 2
    sub_phase: "3b"
    checks:
      - "Technical risks identified with mitigations?"
      - "Dependency risks documented?"
      - "Resource/skill gap risks addressed?"
      - "Contingency plans defined for high risks?"
    scoring:
      excellent: "Comprehensive risk assessment with mitigations"
      acceptable: "Major risks identified, some gaps"
      poor: "Risks not assessed or no mitigations"

  - id: test_strategy
    name: "Test Strategy"
    weight: 2
    sub_phase: "3b"
    checks:
      - "Unit test scope defined?"
      - "Integration test scope defined?"
      - "E2E test scenarios from user stories?"
      - "Coverage target specified (minimum 80%)?"
      - "Test tasks included in breakdown?"
    scoring:
      excellent: "Complete test strategy with coverage targets"
      acceptable: "Basic test strategy, gaps in coverage plan"
      poor: "Test strategy missing or incomplete"

  # ===== PHASE 3c CRITERIA =====
  - id: jira_completeness
    name: "Jira Issue Completeness"
    weight: 3
    sub_phase: "3c"
    checks:
      - "All tasks from breakdown have Jira issues?"
      - "Every issue has full description from consolidated spec?"
      - "Every issue has acceptance criteria?"
      - "Every issue has estimate?"
    scoring:
      excellent: "All issues complete with full information"
      acceptable: "Most issues complete, minor gaps"
      poor: "Issues missing or incomplete"

  - id: jira_information_transfer
    name: "Information Transfer"
    weight: 3
    sub_phase: "3c"
    checks:
      - "Business context from PRD included in descriptions?"
      - "Technical notes from Tech Spec included?"
      - "Spec section references in every issue?"
      - "No information lost from consolidated spec?"
    scoring:
      excellent: "Full information transfer to Jira"
      acceptable: "Most information transferred, minor omissions"
      poor: "Significant information missing from Jira"

  - id: jira_structure
    name: "Jira Structure & Links"
    weight: 2
    sub_phase: "3c"
    checks:
      - "Proper hierarchy (Epic > Story > Task > Subtask)?"
      - "Dependencies linked between issues?"
      - "Issues assigned to correct sprints?"
      - "Labels and components properly set?"
    scoring:
      excellent: "Well-structured Jira with all links"
      acceptable: "Good structure, some linking gaps"
      poor: "Poor hierarchy or missing dependencies"

  # ===== CROSS-PHASE CRITERIA =====
  - id: security_tasks
    name: "Security Tasks"
    weight: 2
    sub_phase: "all"
    checks:
      - "Security implementation tasks included?"
      - "Security review checkpoints planned?"
      - "Penetration testing scheduled if needed?"
      - "Security requirements from consolidated spec covered?"
    scoring:
      excellent: "Security tasks properly planned"
      acceptable: "Basic security tasks present"
      poor: "Security tasks missing from plan"

  - id: documentation_tasks
    name: "Documentation Tasks"
    weight: 1
    sub_phase: "all"
    checks:
      - "API documentation tasks included?"
      - "User documentation planned if needed?"
      - "Architecture decision records planned?"
    scoring:
      excellent: "All documentation tasks planned"
      acceptable: "Basic documentation tasks"
      poor: "Documentation not planned"

# Thresholds by sub-phase
thresholds:
  "3a": 7
  "3b": 7
  "3c": 7
  overall: 7

max_iterations: 3

checkpoints:
  - sub_phase: "3a"
    name: "Consolidation Checkpoint"
    requires_user_approval: true
    gate_criteria:
      - "consolidation_completeness >= 7"
      - "contradiction_resolution >= 7"
      - "consolidation_approval = approved"

  - sub_phase: "3b"
    name: "Task Breakdown Checkpoint"
    requires_user_approval: false
    gate_criteria:
      - "task_quality >= 7"
      - "sprint_planning >= 6"
      - "risk_assessment >= 6"
      - "test_strategy >= 6"

  - sub_phase: "3c"
    name: "Jira Population Checkpoint"
    requires_user_approval: false
    gate_criteria:
      - "jira_completeness >= 7"
      - "jira_information_transfer >= 7"
      - "jira_structure >= 6"

notes:
  borderline: "Review task dependencies, estimation accuracy, and information completeness"
  improvement_focus:
    - "Ensure all contradictions are resolved with rationale"
    - "Ensure tasks are 1-3 hour sized"
    - "Add missing dependencies"
    - "Include all information from consolidated spec in Jira"
    - "Include testing and documentation tasks"

  phase_3a_focus:
    - "No information may be lost during consolidation"
    - "PRD is source of truth unless Research contradicts"
    - "User must approve before proceeding"

  phase_3b_focus:
    - "Tasks should be independently completable"
    - "Every task must link to consolidated spec"
    - "Critical path must be visible"

  phase_3c_focus:
    - "Jira descriptions should be comprehensive"
    - "All acceptance criteria must transfer"
    - "Dependencies must be linked, not just documented"
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