init-pipeline

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates shell scripts (e.g., .claude/hooks/enforce-classification.sh and quality-gate.sh) and makes them executable using chmod +x. These scripts are then registered as Claude Code hooks to provide automated feedback and enforcement during the development process.
  • [EXTERNAL_DOWNLOADS]: The skill detects and recommends standard third-party development tools such as Biome, Lefthook, and pnpm. It uses npx to run the only-allow package manager enforcement tool.
  • [DYNAMIC_EXECUTION]: The skill constructs shell scripts at runtime, incorporating user-defined file patterns (e.g., IMPL_PATTERNS) to customize the classification gate logic for the specific project context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 02:59 PM