hone
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed to audit AI CLI configurations (Codex, Antigravity, Claude Code) and manage lifecycle hooks, with a primary focus on security posture and best practices.
- [SAFE]: It contains explicit prohibitions against reading sensitive files such as authentication tokens, credentials, or session history (~/.claude/credentials.json, ~/.codex/auth.json, etc.).
- [SAFE]: The skill provides advanced security recipes for Claude Code hooks, including patterns to block destructive commands (e.g., rm -rf /), detect secrets in file writes (e.g., API keys, private keys), and enforce least-privilege MCP access.
- [SAFE]: It handles personal environment and dotfile management by advocating for industry-standard tools (Stow, chezmoi, yadm) and security practices like XDG compliance and secret separation.
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it fetches external documentation and processes local configuration files. However, it incorporates mitigation strategies, such as mandatory safety checks on web-fetched content and instruction boundaries that prevent auto-editing of core configurations.
Audit Metadata