epic-motion-diagnostic
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data regarding business metrics (ACV, NRR, TAM) and go-to-market challenges. While this is a theoretical attack surface, the risk is mitigated by the skill's strict internal scoring logic, mandatory assumption flagging, and defensive operating principles defined in the reference files.
- Ingestion points: User answers to GTM diagnostic questions in
SKILL.md. - Boundary markers: Present via
Assumption:prefixes and explicit question budget constraints. - Capability inventory: Utilizes
web_searchandweb_fetchfor factual verification as peroperating-principles.md. - Sanitization: Employs a strict 4-tier source hierarchy for verifying external data and a competitor verification protocol.
- [EXTERNAL_DOWNLOADS]: The skill references the author's own websites for consultation and the EPIC framework attribution. These are identified as legitimate vendor resources ('shashwatgtm.github.io' and 'gtmexpert.com') belonging to Shashwat Ghosh.
Audit Metadata