skill-creator-plus
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a generator for new instruction files (SKILL.md), which creates a surface for indirect prompt injection where user-supplied goals could result in the generation of malicious instructions.
- Ingestion points: User input provided during the skill creation process (e.g., specific goals, triggers, and resources) is used to populate the body of the generated SKILL.md file.
- Boundary markers: The resulting SKILL.md template does not include explicit boundary markers to separate the generated workflow instructions from potentially untrusted user-supplied content.
- Capability inventory: The skill utilizes local scripts (scripts/init_skill.py, scripts/package_skill.py, scripts/quick_validate.py) that perform file system operations, including directory creation, file writing, and zip archive generation.
- Sanitization: The skill's scripts implement robust validation for technical metadata, such as regex checks for skill names in init_skill.py and extensive path validation in package_skill.py to prevent path traversal attacks, although the generated natural language instructions are not sanitized.
Audit Metadata