signals-scout-web-analytics
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill inherently processes untrusted data originating from web browsers (URLs, referrer domains, and UTM parameters) which are ingested via the
eventsandsessionstables. - Ingestion points: Web analytics data is queried from the
eventsandsessionstables in SKILL.md. - Boundary markers: The instructions explicitly direct the agent to quote untrusted data as short snippets and avoid treating values as instructions.
- Capability inventory: The skill uses
execute-sqlfor analysis andemit_report/edit_reportto output findings, as well asscout-scratchpad-rememberfor internal state. - Sanitization: The skill mandates aggressive truncation of user-supplied strings and the use of sanitized identifiers (slugified paths/domains) when storing data in memory.
- [SAFE]: The skill demonstrates high security awareness by including a dedicated 'Untrusted data' section. This section provides proactive remediation guidance to prevent the agent from being influenced by malicious payloads embedded in analytics data (referrer spam), correctly identifying it as an 'attacker-adjacent' stream.
Audit Metadata