clean-functions
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No security vulnerabilities were identified. The skill is limited to providing text-based guidelines and examples for Python function refactoring.
- [NO_CODE]: The skill does not contain any executable scripts or binaries. Its functionality is implemented through Markdown instructions for the AI agent.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze and potentially delete user-provided code based on clean code principles. While this creates a surface where malicious input could influence agent actions, the risk is inherent to the skill's purpose as a code assistant and is handled by base agent safety layers.
- Ingestion points: Python functions provided in the agent's conversation context as described in the frontmatter.
- Boundary markers: None specific to this skill; it relies on the host agent's context boundaries.
- Capability inventory: The skill itself contains no scripts or tools, but guides the agent's native ability to edit files.
- Sanitization: No specific sanitization or filtering of input code is performed by the skill.
Audit Metadata