integration-listing-optimization
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that requires the agent to ingest and analyze untrusted external data sources, specifically customer support tickets and sales-call notes, to perform keyword extraction and research.
- Ingestion points: SKILL.md (Interview Question 4) explicitly prompts the user to provide text from support tickets and sales notes.
- Boundary markers: The instructions do not provide explicit boundary markers or instructions to disregard potential malicious directives embedded within the ingested data.
- Capability inventory: The skill itself has a low-risk profile as it lacks dangerous capabilities; there are no network operations, file-system writes (aside from a suggested local context file), or subprocess/command execution patterns.
- Sanitization: No sanitization or validation steps are defined for the input data derived from these external sources.
Audit Metadata