verify
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting untrusted data from the source code files it is tasked to verify, as well as output from external security scanners. This creates a surface where malicious content within the code could attempt to manipulate the verification process or the AI's logic.
- Ingestion points: Reads code files at specific finding locations, fix records (diffs), and scanner execution output (Step 2 and Step 3).
- Boundary markers: The instructions do not specify the use of delimiters or clear boundaries to isolate the external code content from the agent's internal reasoning instructions.
- Capability inventory: The skill uses a 'Read' tool to access local files and has the authority to write updated finding data to
.appsec/findings.jsonand.appsec/fixed-history.json(Step 6). - Sanitization: There is no mention of sanitizing or escaping the content of the files or scanner results before they are processed by the AI context.
Audit Metadata