playbook-design
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill serves as an entry point for user-defined workflow instructions. It incorporates mitigation strategies such as quality checks for vague success criteria and structured YAML output to maintain data/instruction separation for downstream processing.
- [COMMAND_EXECUTION]: The skill permits users to define shell commands in the configuration artifact. These commands are stored as static strings for future use and are not executed during the design or saving phases.
- [DATA_EXPOSURE]: File system activity is limited to the
./playbooks/path. The use of kebab-case slugs for playbook names prevents directory traversal and unauthorized file access.
Audit Metadata