How navruzbek05/pre-launch-audit-skill works
This repository packages a Claude Code skill for pre-launch audits, not an audit application or scanner library. The documented workflow is: a user requests an audit, Claude Code follows the skill instructions to identify the target stack and inspect the project against nine readiness checks, then saves an evidence-backed report and summarizes it in chat. The README identifies SKILL.md as the main skill and references/search-patterns.md as stack-specific search guidance. Source excerpts were unavailable, so the graph reflects documented relationships only.
Main entry points: SKILL.md Audit instructions · search-patterns.md Stack search patterns
Components 8
Audit workflow
- Claude Code
- Audit instructions
SKILL.md - Nine readiness checks
SKILL.md - Audit report
- Chat summary
Search guidance
- Stack search patterns
references/search-patterns.md
Other components
- User
- Target codebase
Connections 7
- User → Claude Code: requests audit
- Claude Code → Audit instructions: follows instructions
- Claude Code → Target codebase: scans code
- Claude Code → Stack search patterns: uses search guidance
- Audit instructions → Nine readiness checks: defines checks
- Claude Code → Audit report: writes evidence
- Claude Code → Chat summary: presents results