app-security-gate
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by instructing the agent to 'Inspect the repository' and implementation details of the application being audited (SKILL.md, Section: Process).
- [INDIRECT_PROMPT_INJECTION]: The skill possesses high capabilities including the ability to perform 'Safe Auto-Fixes' such as writing validation schemas, modifying database migrations, and changing security headers (SKILL.md, Section: Safe Auto-Fix Policy).
- [INDIRECT_PROMPT_INJECTION]: There are no specific boundary markers or 'ignore embedded instructions' warnings mentioned to protect the agent when it processes potentially adversarial content within the target repository's code or documentation.
- [INDIRECT_PROMPT_INJECTION]: The lack of explicit sanitization or validation of the untrusted code being analyzed, combined with the power to modify that same codebase, creates a surface where an attacker could embed instructions in a target repo to manipulate the auditor agent.
Audit Metadata