socratic-review
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code (snippets, diffs, or file paths) provided by the user, which creates a surface for indirect prompt injection attacks where malicious instructions could be embedded in the code being reviewed.
- Ingestion points: The skill accepts a 'file path, code snippet, diff or SHA' as input for assessment as defined in the frontmatter of
SKILL.md. - Boundary markers: No explicit delimiters or specific 'ignore' instructions are defined to prevent the agent or subagents from following instructions embedded within the reviewed code content.
- Capability inventory: The skill utilizes the
Agenttool to dispatch subagents and performs file system traversal to trace collaborators and call sites across the project. - Sanitization: The instructions do not implement sanitization or escaping of the user-provided code before it is analyzed by the AI.
Audit Metadata