金大哥 - gap-analysis-workflow Skill 详情

Intent

Run the composed workflow gap-analysis-workflow using atomic capability skills to systematically identify what capabilities are missing and prioritize their development.

Success criteria:

Current capability coverage mapped with evidence
Gaps identified with clear justification
Relationships between existing and missing capabilities documented
Prioritized roadmap for new skill development
Audit trail of analysis process

Compatible schemas:

reference/capability_ontology.yaml
reference/workflow_catalog.yaml

Inputs

| Parameter | Required | Type | Description | |-----------|----------|------|-------------| | goal | Yes | string | The analysis objective (e.g., "identify gaps for autonomous deployment") | | scope | Yes | string|array | Domain, layer, or capability set to analyze | | constraints | No | object | Limits (e.g., max new skills, priority criteria, timeline) | | reference_ontology | No | string | Path to reference ontology for comparison | | existing_capabilities | No | array | List of already-implemented capabilities |

Procedure

Create checkpoint marker if mutation might occur:
- Create .claude/checkpoint.ok after confirming rollback strategy
Invoke /inspect and store output as inspect_out
- Examine current capability landscape and documentation
Invoke /map-relationships and store output as map-relationships_out
- Map dependencies and connections between existing capabilities
Invoke /discover-relationship and store output as discover-relationship_out
- Identify implicit relationships and missing links
Invoke /compare-plans and store output as compare-plans_out
- Compare current state against ideal or reference ontology
Invoke /prioritize and store output as prioritize_out
- Rank gaps by impact, effort, and strategic value
Invoke /generate-plan and store output as generate-plan_out
- Create development roadmap for new capabilities
Invoke /audit and store output as audit_out
- Record analysis process and evidence

Output Contract

Return a structured object:

workflow_id: string  # Unique analysis execution ID
goal: string  # Analysis objective
status: completed | partial | failed
current_state:
  capabilities_analyzed: integer
  coverage_percentage: number  # 0.0-1.0
  layers_covered: array[string]
  evidence_anchors: array[string]
gaps_identified:
  total: integer
  by_layer:
    perception: array[string]
    modeling: array[string]
    reasoning: array[string]
    action: array[string]
    safety: array[string]
    meta: array[string]
  by_priority:
    critical: array[string]
    high: array[string]
    medium: array[string]
    low: array[string]
  evidence_anchors: array[string]
relationships:
  existing_dependencies: array[object]
  missing_connections: array[object]
  orphan_capabilities: array[string]
  evidence_anchors: array[string]
comparison:
  reference_ontology: string
  alignment_score: number  # 0.0-1.0
  divergences: array[string]
  evidence_anchors: array[string]
roadmap:
  phases: array[object]
  total_new_skills: integer
  estimated_effort: string
  dependencies: array[object]
  evidence_anchors: array[string]
audit:
  log_path: string
  methodology: string
  evidence_anchors: array[string]
confidence: number  # 0.0-1.0
evidence_anchors: array[string]
assumptions: array[string]

Field Definitions

| Field | Type | Description | |-------|------|-------------| | workflow_id | string | Unique identifier for this analysis | | current_state | object | Summary of existing capability coverage | | gaps_identified | object | Missing capabilities organized by layer and priority | | relationships | object | Dependency mapping including missing connections | | comparison | object | Alignment with reference ontology | | roadmap | object | Phased development plan for new skills | | audit | object | Analysis methodology and evidence trail | | confidence | number | 0.0-1.0 based on evidence completeness | | evidence_anchors | array | All evidence references collected | | assumptions | array | Explicit assumptions made during analysis |

Examples

Example 1: Gap Analysis for Autonomous Code Review

Input:

goal: "Identify gaps for fully autonomous code review capability"
scope:
  - "reasoning"
  - "safety"
  - "action"
constraints:
  max_new_skills: 10
  priority_criteria:
    - "security_impact"
    - "automation_potential"
reference_ontology: "schemas/capability_ontology.yaml"

Output:

workflow_id: "gap_20240115_120000_codereview"
goal: "Identify gaps for fully autonomous code review capability"
status: completed
current_state:
  capabilities_analyzed: 45
  coverage_percentage: 0.72
  layers_covered:
    - "reasoning"
    - "safety"
    - "action"
  evidence_anchors:
    - "file:schemas/capability_ontology.yaml"
    - "file:skills/critique/SKILL.md"
gaps_identified:
  total: 8
  by_layer:
    perception: []
    modeling:
      - "detect-code-smell"
      - "identify-security-pattern"
    reasoning:
      - "compare-implementations"
      - "evaluate-test-coverage"
    action:
      - "generate-review-comment"
      - "apply-suggested-fix"
    safety:
      - "verify-no-regression"
      - "constrain-auto-merge"
    meta: []
  by_priority:
    critical:
      - "verify-no-regression"
      - "identify-security-pattern"
    high:
      - "detect-code-smell"
      - "constrain-auto-merge"
    medium:
      - "compare-implementations"
      - "generate-review-comment"
    low:
      - "evaluate-test-coverage"
      - "apply-suggested-fix"
  evidence_anchors:
    - "file:schemas/capability_ontology.yaml:nodes"
    - "tool:compare-plans:coverage_analysis"
relationships:
  existing_dependencies:
    - from: "critique"
      to: "evaluate"
      type: "requires"
    - from: "plan"
      to: "critique"
      type: "soft_requires"
  missing_connections:
    - from: "detect-code-smell"
      to: "critique"
      type: "should_precede"
      reason: "Code smells inform critique priorities"
    - from: "verify-no-regression"
      to: "act-plan"
      type: "must_follow"
      reason: "Regression check required after any code change"
  orphan_capabilities: []
  evidence_anchors:
    - "tool:map-relationships:dependency_graph"
    - "tool:discover-relationship:implicit_links"
comparison:
  reference_ontology: "schemas/capability_ontology.yaml"
  alignment_score: 0.72
  divergences:
    - "Missing specialized detection capabilities for code patterns"
    - "No automated fix application in action layer"
    - "Regression verification not formalized"
  evidence_anchors:
    - "tool:compare-plans:ontology_diff"
roadmap:
  phases:
    - phase: 1
      name: "Security Foundation"
      skills:
        - "identify-security-pattern"
        - "verify-no-regression"
      rationale: "Critical for safe autonomous operation"
    - phase: 2
      name: "Detection Enhancement"
      skills:
        - "detect-code-smell"
        - "constrain-auto-merge"
      rationale: "Improves review quality and safety"
    - phase: 3
      name: "Automation Expansion"
      skills:
        - "compare-implementations"
        - "generate-review-comment"
        - "evaluate-test-coverage"
        - "apply-suggested-fix"
      rationale: "Full autonomous review capability"
  total_new_skills: 8
  estimated_effort: "2-3 sprints"
  dependencies:
    - skill: "apply-suggested-fix"
      requires: ["verify-no-regression", "constrain-auto-merge"]
  evidence_anchors:
    - "tool:prioritize:impact_matrix"
    - "tool:generate-plan:roadmap"
audit:
  log_path: ".claude/audit/gap_20240115_120000_codereview.log"
  methodology: "Systematic comparison against reference ontology with layer-by-layer analysis"
  evidence_anchors:
    - "file:.claude/audit/gap_20240115_120000_codereview.log"
confidence: 0.85
evidence_anchors:
  - "file:schemas/capability_ontology.yaml"
  - "tool:map-relationships:dependency_graph"
  - "tool:compare-plans:ontology_diff"
  - "tool:prioritize:impact_matrix"
assumptions:
  - "Reference ontology is current and complete"
  - "Existing skills are correctly implemented"
  - "Priority criteria reflect actual business needs"

Evidence pattern: Ontology comparison, dependency graph analysis, impact-based prioritization.

Verification

[ ] Coverage Analyzed: All capabilities in scope examined
[ ] Gaps Documented: Each gap has layer classification and priority
[ ] Relationships Mapped: Dependencies and missing connections identified
[ ] Comparison Complete: Alignment score computed against reference
[ ] Roadmap Generated: Phased plan with dependencies
[ ] Audit Trail: Analysis methodology documented

Verification tools: Read (for ontology files), Grep (for capability search), Bash (for validation)

Safety Constraints

mutation: false
requires_checkpoint: false
requires_approval: false
risk: low

Capability-specific rules:

Do not create new skills during analysis (discovery only)
Validate all capability references against ontology
Document assumptions about missing capabilities
Flag potential security implications of gaps
Preserve existing ontology structure

Composition Patterns

Commonly follows:

inspect - After initial codebase exploration
retrieve - After fetching reference documentation

Commonly precedes:

generate-plan - To create detailed skill specifications
prioritize - To refine gap prioritization
summarize - To create executive summary

Anti-patterns:

Never skip relationship mapping before prioritization
Never propose skills without checking ontology for existing alternatives
Never prioritize without defined criteria
Never generate roadmap without dependency analysis

Workflow references:

See reference/workflow_catalog.yaml#gap-analysis-workflow for step definitions
See reference/capability_ontology.yaml for reference structure