skill-creator
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/init_skill.pymodifies file system permissions on dynamically created content. \n - Evidence: The script uses
example_script.chmod(0o755)to set executable permissions on a newly generated template script (example.py). \n- [PROMPT_INJECTION]: The skill facilitates the generation of instructions and structures based on user-supplied inputs, creating a surface for potential indirect prompt injection. \n - Ingestion points: The
skill_nameandpathparameters are accepted as command-line arguments inscripts/init_skill.py. \n - Boundary markers: No specific delimiters or boundary warnings are used to wrap user-provided content in the generated files. \n
- Capability inventory: The skill has the capability to create directories (
Path.mkdir), write files (Path.write_text), and modify permissions (chmod). \n - Sanitization: Basic regex validation is performed on the skill name in
scripts/quick_validate.pyto ensure it follows hyphen-case conventions.
Audit Metadata