signals-scout-surveys
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill aggregates open-text survey responses to generate reports, which involves processing untrusted user-supplied content.
- Ingestion points: Untrusted data enters the agent context via
execute-sqlqueries that retrieve survey response properties fromsurvey sentevents, as described inreferences/response-querying.md. - Boundary markers: The instructions do not mandate specific delimiters or safety headers when the agent processes the retrieved survey text.
- Capability inventory: The skill can execute shell commands via MCP tools, modify project scratchpads, and publish/edit reports using
scout-emit-reportandscout-edit-report. - Sanitization: The instructions include robust sanitization guidelines, such as filtering internal test responses, stripping PII, and truncating responses to the sentence level, which serve as partial mitigations against malicious payloads.
Audit Metadata