spec-kit-analyze

Perform a read-only consistency analysis across all available change artifacts.

Input: Optionally specify a change name. If omitted, infer from context or ask.

This skill makes no file modifications.

Steps

1. Load artifacts

   Load what exists under specs/changes/<name>/:

   • spec.md (required)
   • plan.md (if present)
   • tasks.md (if present)

   Also load FRAMEWORK.md as the governing principles baseline.

2. Run consistency checks

Analyze across these dimensions:

Dimension	What to check
Coverage	Every user story in spec.md has corresponding tasks in tasks.md
Alignment	Plan architecture matches spec goals and constraints
FRAMEWORK.md compliance	No dependency rule violations, no anemic domain model, ports present for external dependencies
Duplication	Overlapping requirements or tasks
Ambiguity	Vague terms without measurable criteria
Terminology drift	Same concept named differently across artifacts
Orphans	Tasks with no corresponding requirement, or requirements with no tasks

1. Produce findings

   Output a structured findings table:

   ## Analysis: <change-name>
   
   ### Findings
   | ID | Severity | Dimension | Description | Location |
   |----|----------|-----------|-------------|----------|
   
   ### Coverage Summary
   | Requirement | Task IDs |
   |-------------|----------|
   
   ### FRAMEWORK.md Violations
   <List any violations. CRITICAL severity — must be fixed before proceeding.>
   
   ### Metrics
   - Total findings: N
   - Critical: N | High: N | Medium: N | Low: N
   
   ### Recommended Next Actions
   <Ordered list of what to fix.>
   

   Severity: CRITICAL (FRAMEWORK.md violation) > HIGH > MEDIUM > LOW.

Output

After the analysis:

• If CRITICAL findings exist: "Fix FRAMEWORK.md violations before proceeding. Do not use the spec-kit-tasks or spec-kit-implement skills until these are resolved."
• If no critical findings: "Use the spec-kit-tasks skill to generate the implementation task list."

Guardrails

• Read-only. Never modify any file.
• FRAMEWORK.md violations are always CRITICAL — never downgrade or ignore them.
• Deterministic: re-running on the same artifacts produces the same findings.