ship-check
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted input from external pull request components, which creates a potential surface for indirect prompt injection.
- Ingestion points: The skill ingests data from pull-request descriptions, comments, review text, issue text, CI logs, and test fixtures (SKILL.md).
- Boundary markers: There are no explicit instructions to use delimiters or markers to isolate untrusted data during processing, though the skill does warn the agent that these sources are untrusted.
- Capability inventory: The skill has access to repository contents, CI logs, and the ability to execute commands for behavioral verification (SKILL.md).
- Sanitization: No specific sanitization or filtering procedures are mandated for the external data before it is handled by the agent.
Audit Metadata