signals-scout-revenue-analytics
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting untrusted data from external sources and using it to generate human-readable reports.
- Ingestion points: Untrusted data enters the agent's context through several tools described in SKILL.md, including project-get (which retrieves user-defined event names), external-data-sync-logs (which retrieves error strings from external payment platforms like Stripe), and execute-sql (which reads records from database views).
- Boundary markers: The instructions do not specify the use of delimiters (e.g., XML tags or triple quotes) or 'ignore' instructions when interpolating these external strings into the report templates for scout-emit-report.
- Capability inventory: The skill possesses write capabilities, specifically the ability to author and edit reports in the PostHog inbox via scout-emit-report and scout-edit-report.
- Sanitization: There is no evidence of sanitization, escaping, or validation of the external content (like error messages) before it is included in the output report body.
Audit Metadata