guidewire-core-workflow-a
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data returned from the Guidewire Cloud API. This external data acts as an ingestion point that could potentially influence agent behavior if the API response contains malicious instructions. While this is a common attack surface for API-driven skills, the lack of explicit boundary markers is noted.
- Ingestion points: Responses from
${BASE}/pc/rest/v1/endpoints processed inSKILL.mdandreferences/implementation-guide.md. - Boundary markers: The skill does not define specific delimiters to isolate API-provided data from agent instructions.
- Capability inventory: The agent has access to
Bash(curl:*),Bash(jq:*),Read,Write, andEdittools. - Sanitization: No explicit sanitization logic is provided for strings retrieved from the API before they are utilized in the workflow.
Audit Metadata