execute-automated-fix
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the
issuecontext variable, which includes user-provided descriptions and reproduction steps. - Ingestion points: The
issuevariable is the primary source of untrusted data processed by the agent inSKILL.md. - Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the issue content as data rather than instructions.
- Capability inventory: The agent has the capability to modify the project's filesystem ("Fix the issue") and likely execute shell commands or scripts to "Reproduce the issue" and "Verify the fix".
- Sanitization: There is no evidence of sanitization or filtering of the
issuecontent before it is used by the agent.
Audit Metadata