eli5
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill consumes untrusted data by reading external artifacts and code diffs to generate summaries.
- Ingestion points: Workflow step 1 in SKILL.md requires reading the "complete referenced artifact" and session diffs.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between its internal instructions and the content of the analyzed artifacts.
- Capability inventory: The agent utilizes file system tools and git commands to access the data required for analysis.
- Sanitization: There is no requirement or mechanism specified for sanitizing input data before processing.
Audit Metadata