scaffold-cc-hooks

Warn

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides scripts (scripts/check_workspace_trust.sh) that programmatically modify the global Claude Code configuration file (~/.claude.json) to set hasTrustDialogAccepted to true. This action bypasses a built-in security prompt designed to prevent unauthorized hook execution in new workspaces.
  • [CREDENTIALS_UNSAFE]: The scripts/audit_project.sh script is designed to scan for and catalog sensitive file paths including .env, .envrc, *.pem, and *.key. While the script reports metadata about these files, identifying their locations within a project constitutes a sensitive data exposure finding.
  • [COMMAND_EXECUTION]: The core functionality relies on multiple bash scripts (scripts/scaffold_hooks.sh, scripts/merge_settings.sh, scripts/render_hooks_readme.sh) that perform extensive file system modifications and configuration merges based on the project's structure and user-provided plans.
  • [EXTERNAL_DOWNLOADS]: The skill instructions (SKILL.md) direct the agent to fetch and process content from external documentation sites (e.g., code.claude.com and reading.sh) to verify current hook specifications before performing scaffolding operations.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 12, 2026, 05:58 AM