Vcsdd spec review
Run Phase 1c (spec review gate) for the active VCSDD feature. Spawns a fresh vcsdd-adversary instance to review the behavioral spec and verification architecture. Records PASS/FAIL gate verdict.From its SKILL.md
npx -y skills add sc30gsw/vcsdd-claude-code --skill vcsdd-spec-reviewAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
SKILL.md
1.6 KB, as published. Nobody here has run it
What
Runs the spec review gate (Phase 1c). Spawns a fresh vcsdd-adversary agent to review the behavioral specification and verification architecture for completeness, correctness, and verification readiness. Records the gate verdict.
When
Run after /vcsdd-spec completes Phase 1a and 1b. Requires active feature at phase 1b.
How
- Write spec review manifest to
reviews/spec/iteration-N/input/manifest.json:{ "reviewType": "spec", "artifactsToReview": ["specs/behavioral-spec.md", "specs/verification-architecture.md"], "reviewDimensions": ["spec_fidelity", "verification_readiness"] } - Spawn fresh vcsdd-adversary agent (new context, no Builder history)
- Adversary reads and reviews spec files only (no source code yet)
- Collect verdict from
reviews/spec/iteration-N/output/verdict.json - Record adversary verdict:
recordGate(feature, '1c', verdict, 'adversary') - If PASS in lean mode: transition to Phase 2a when the user is ready
- If PASS in strict mode: require explicit user confirmation, then record human approval:
recordGate(feature, '1c', 'PASS', 'human', { approvedBasedOn: 'adversary' }) - If FAIL: display findings, check iteration limit (max 3), prompt for revision
Examples
/vcsdd-spec-review
/vcsdd-spec-review --auto-approve # strict mode: record the human approval immediately after a PASS