refactor-guide
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted code snippets provided by the user, which could contain hidden instructions designed to manipulate the agent's behavior.
- Ingestion points: Code snippets provided by the user during the workflow steps defined in
SKILL.md. - Boundary markers: The instructions do not specify delimiters or explicit guidelines to ignore instructions that might be embedded within the provided code samples.
- Capability inventory: The skill is instructed to write user context into a local memory file,
SKILL_MEMORY.md(SKILL.md, Step 1). - Sanitization: No sanitization or validation logic is defined for the input code before it is analyzed by the agent or recorded to the memory file.
- [NO_CODE]: This skill consists of instructional Markdown and reference documentation only. It does not include or execute any script files (e.g., Python, JavaScript, or Shell scripts).
Audit Metadata