speckit-docguard-review
Diagnose documentation issues and generate AI-ready fix prompts
Files1
View on GitHubSKILL.md.claude/skills/speckit-docguard-review/SKILL.md
GitHubDocGuard Diagnose
Runs guard validation, analyzes failures, and generates AI-ready prompts to fix documentation issues.
User Input
$ARGUMENTS
Steps
- Run DocGuard diagnose on the current project:
npx --yes docguard-cli@latest diagnose $ARGUMENTS
Review the output:
- Issues found — categorized as errors or warnings
- Remediation plan — ordered list of fix commands
- AI-ready prompt — copy/paste to your AI agent for automated fixes
For multi-perspective analysis, add
--debateto get three-agent prompts (Advocate/Challenger/Synthesizer).
Flags
--format json— Output as JSON--debate— Generate multi-perspective AI prompts--dir <path>— Run on a different directory
Research
Inspired by AITPG (IEEE TSE 2026) multi-agent prompting and TRACE (IEEE TMLCN 2026) calibrated quality evaluation.