Claude Code Plugins

Community-maintained marketplace

Feedback

Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name speckit-analyze
description Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
allowed-tools Bash, Read, Grep, Glob

Spec-Kit Analyze

Cross-artifact consistency and quality analysis for feature specifications. Validation step after creating tasks.md.

When to Use

  • After creating tasks.md with speckit-tasks
  • Before starting implementation with speckit-implement
  • Need to identify gaps or conflicts before coding

Execution Workflow

STRICTLY READ-ONLY - Does not modify any files.

  1. Initialize analysis context: Run .specify/scripts/bash/check-prerequisites.sh --json --require-tasks --include-tasks to get FEATURE_DIR and AVAILABLE_DOCS. Derive paths:
    • SPEC = FEATURE_DIR/spec.md
    • PLAN = FEATURE_DIR/plan.md
    • TASKS = FEATURE_DIR/tasks.md
    • CONSTITUTION = .specify/memory/constitution.md
  2. Load artifacts (progressive disclosure - minimal context):
    • From spec.md: Overview, Functional Requirements, Non-Functional Requirements, User Stories, Edge Cases
    • From plan.md: Architecture/stack choices, Data Model references, Phases, Technical constraints
    • From tasks.md: Task IDs, Descriptions, Phase grouping, Parallel markers [P], File paths
    • From constitution.md: Project principles, MUST/SHOULD requirements
  3. Build semantic models (internal representations):
    • Requirements inventory with stable keys
    • User story/action inventory with acceptance criteria
    • Task coverage mapping (maps tasks to requirements)
    • Constitution rule set (principle names and normative statements)
  4. Detection passes (maximum 50 findings total - aggregate overflow):
    • Duplication: Near-duplicate requirements, lower-quality phrasing
    • Ambiguity: Vague adjectives (fast, scalable, secure, intuitive, robust), unresolved placeholders (TODO, ???, <placeholder>)
    • Underspecification: Requirements missing object/outcome, user stories missing acceptance criteria, tasks referencing undefined files
    • Constitution Alignment: Conflicts with MUST principles (ALWAYS CRITICAL), missing mandated sections/gates
    • Coverage Gaps: Requirements with zero tasks, tasks with no mapped requirement, missing non-functional task coverage
    • Inconsistency: Terminology drift, data entities in plan but not spec, task ordering contradictions, conflicting requirements
  5. Severity assignment:
    • CRITICAL: Violates constitution MUST, missing core spec artifact, requirement with zero coverage blocking baseline
    • HIGH: Duplicate or conflicting requirement, ambiguous security/performance attribute, untestable acceptance criterion
    • MEDIUM: Terminology drift, missing non-functional task coverage, underspecified edge case
    • LOW: Style/wording improvements, minor redundancy
  6. Produce compact analysis report (Markdown, no file writes):
    • Findings table: ID, Category, Severity, Location(s), Summary, Recommendation
    • Coverage summary: Requirement → Tasks mapping
    • Constitution alignment issues table
    • Unmapped tasks table
    • Metrics: Total requirements, total tasks, coverage %, ambiguity count, duplication count, critical/high/medium/low issues
  7. Next actions:
    • If CRITICAL: Must resolve before implementation
    • If only LOW/MEDIUM: May proceed with improvements recommended
    • Provide explicit command suggestions
  8. Offer remediation: Ask if user wants concrete edit suggestions (not applied automatically)

Key Points

  • NEVER modify files - read-only analysis
  • Constitution violations are ALWAYS CRITICAL - non-negotiable
  • Progressive disclosure - load artifacts incrementally, don't dump all content
  • Token-efficient output - tables over prose, limit to 50 findings, aggregate overflow
  • Deterministic results - rerunning without changes produces consistent IDs and counts
  • Coverage % guidelines:
    • 100%: Perfect (rare)
    • 80-99%: Excellent
    • 60-79%: Good (review gaps)
    • < 60%: Poor (significant gaps)
  • Report zero issues gracefully - emit success report with coverage statistics

Next Steps

After analysis:

  • Fix CRITICAL - Update spec/plan/tasks
  • Start implementation - Use speckit-implement
  • Re-analyze - After fixes to verify

See Also

  • speckit-plan - Create technical implementation strategy
  • speckit-tasks - Break plan into actionable tasks
  • speckit-implement - Execute implementation plan