meta-reviewing-api-reviewing
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code diffs and server logic, creating a surface for indirect prompt injection. A malicious actor could include hidden instructions in pull requests to manipulate the agent's security review results.\n
- Ingestion points: Analyzes request data, route handlers, and code diffs (SKILL.md).\n
- Boundary markers: Lacks explicit delimiters or instructions to ignore embedded commands in the input data.\n
- Capability inventory: Accesses file content and evaluates programming logic for security flaws.\n
- Sanitization: No sanitization of the input diff content is specified.\n- [EXTERNAL_DOWNLOADS]: The skill references a metadata schema from the author's official GitHub repository.\n
- Evidence:
https://raw.githubusercontent.com/agents-inc/agents-inc/main/packages/cli/src/schemas/metadata.schema.jsonin metadata.yaml.
Audit Metadata