generate-skill
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The ADVANCED.md file includes setup script patterns that utilize 'npm install' for dependency management. While these are placeholders for developers, they represent a surface for downloading external packages.
- [COMMAND_EXECUTION] (LOW): Both ADVANCED.md and PATTERNS.md contain boilerplate shell script templates for task execution, environment detection, and setup. These examples involve the use of system commands.
- [PROMPT_INJECTION] (LOW): Category 8: Indirect Prompt Injection surface. The 'Quick Generation Flow' in ADVANCED.md describes a process of ingesting untrusted user input to generate skill content. Evidence Chain: \n1. Ingestion points: User input collected during Step 1 of the 'Quick Generation Flow'. \n2. Boundary markers: Absent in the provided templates. \n3. Capability inventory: Shell execution, file operations, and dependency installation (npm install). \n4. Sanitization: No sanitization or validation steps are defined in the workflow description.
Audit Metadata