marketing-analytics
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines processes for ingesting and analyzing untrusted external data, such as customer survey responses, interview transcripts, and third-party enrichment data.
- Ingestion points: Primarily located in
agents/analytics-customer-insights-researcher.md(research data) andagents/analytics-gtm-data-strategist.md(external data enrichment). - Capability inventory: The skill is restricted to analytical and documentation tasks; no executable capabilities like shell access or arbitrary network operations are present.
- Sanitization: The
agents/analytics-marketing-ai-ops-lead.mdagent explicitly mandates defensive measures against prompt injection, including 'eval sets' and human-in-the-loop validation for high-risk data flows. - [SAFE]: Analysis of the skill's instructions, agent roles, and referenced external sources (e.g., GDPR documentation, NIST frameworks, and established open-source projects) indicates a standard professional implementation. No obfuscation, credential harvesting, or unauthorized persistence mechanisms were identified.
Audit Metadata