audit-spec
Audit specification documents for ambiguity, consistency, and architectural compliance.
Install
mkdir -p .claude/skills/audit-spec && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/15268" && unzip -o skill.zip -d .claude/skills/audit-spec && rm skill.zipInstalls to .claude/skills/audit-spec
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Audit specification documents for ambiguity, consistency, and architectural compliance.About this skill
Skill: Spec Linter (Specification Audit)
<role_gate>
<required_agent>QualityGuard</required_agent>
<instruction>
Before proceeding with any instructions, you MUST strictly check that your ACTIVE_AGENT_ID matches the required_agent above.
Match Case:
- Proceed normally.
Mismatch Case:
- You MUST read the file
.github/agents/{required_agent}.agent.md. - You MUST ADOPT the persona defined in that file for the duration of this skill.
- Proceed with the skill acting as the {required_agent}.
You are supporting the @QualityGuard. Your goal is to review specification documents (Requirements, Design, Plans) to ensure they are clear, consistent, and compliant with project standards before implementation begins.
📋 Task Initialization
IMMEDIATELY use the #todo tool to register the following tasks:
- Ambiguity Check: Scan for vague language.
- Consistency Check: Verify alignment between strict requirements and technical design.
- Architecture Check: Ensure compliance with
AGENTS.mdrules. - Generate Report: Output the Spec Audit Report.
🎯 Objective
Review the provided specification document(s) to prevent "Garbage In, Garbage Out". Ensure the developer has a crystal-clear, feasible, and compliant plan.
🛡️ Audit Steps
1. Ambiguity Check
- Scan for Vague Words: Look for phrases like:
- "maybe", "probably", "roughly", "about"
- "should be able to" (unless strictly defined)
- "user friendly", "fast" (without metrics)
- "etc.", "and so on"
- Action: Flag these as warnings requiring clarification.
2. Consistency Check
- Story vs. Design: Ensure every User Story has a corresponding technical plan (API endpoint, DB schema change, or UI component).
- Missing Technical Details: If a story requires a new field, is the DB migration specified? If it implies a new page, is the route defined?
3. Architecture Compliance (AGENTS.md)
- Check for Violations:
- Does the design use the Repository Pattern? (Required)
- Are Agents incorrectly assigned runtime roles? (Agents are build-time only).
- Are strict rules from
AGENTS.mdfollowed?
📝 Output Format
You must output a Spec Audit Report.
# Spec Audit Report
## 1. Summary
[Brief summary of the spec quality. Is it ready for implementation?]
## 2. Issues & discrepancies
### 🔴 Critical (Must Fix)
- [ ] **Consistency**: User Story "X" implies a database change, but no schema update is listed in the Design section.
- [ ] **Architecture**: The design proposes a direct database query in the controller, violating the Repository Pattern.
### 🟡 Warnings (Clarification Needed)
- [ ] **Ambiguity**: "The system should be fast" - Define "fast" (e.g., < 200ms).
- [ ] **Ambiguity**: "Handle various errors" - List specific error codes to handle.
## 3. Conclusion
[ ] **APPROVE**: The spec is clear, consistent, and compliant. Proceed to Implementation.
[ ] **REQUEST CHANGES**: Address the critical issues above before proceeding.