prompt-factory
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The requirements.txt file confirms the skill relies solely on the Python standard library. No external dependencies or remote scripts are downloaded during execution.
- [COMMAND_EXECUTION]: The skill includes Python scripts like generate_prompt.py and batch_generator.py for local automation. These scripts are restricted to basic file I/O and string processing, with no evidence of arbitrary system command execution or unsafe subprocess calls.
- [DATA_EXFILTRATION]: No network-capable modules (e.g., requests, curl) are used in the automation scripts. Processing is entirely local to the environment, and no sensitive file paths or credentials are accessed.
- [PROMPT_INJECTION]: The skill's instructions in SKILL.md include clear constraints that prevent the agent from executing the content of the generated prompts. This isolation ensures that the output is treated strictly as a deliverable for the user.
- [SAFE]: The skill is authored by 'ckorhonen' and exhibits consistent, professional behavior aligned with its stated purpose of prompt generation. It follows secure coding practices and maintains a narrow operational scope.
Audit Metadata