migrating-vercel-ai-sdk-and-eve-to-pydantic-ai
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze untrusted content from external repositories, including source code, tests, and configuration files.
- Ingestion points: Processes repository source code, dependency manifests, lockfiles, and instruction files as specified in
SKILL.mdandreferences/VERIFICATION-AND-CUTOVER.md. - Boundary markers: The instructions do not provide specific delimiters or ignore-content directives to ensure the agent does not execute instructions embedded in the analyzed code.
- Capability inventory: The agent is tasked with deep logic analysis and code generation, which necessitates following instructions found within the processed files.
- Sanitization: No explicit sanitization logic or content validation is defined for the agent's ingestion of repository data.
Audit Metadata