sesion-mental-health-saas
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to clone an external project repository from GitHub (github.com/fahad-hamid/psique-workflow-clinic.git). GitHub is a well-known service and the repository serves as the foundation for the platform's codebase.
- [PROMPT_INJECTION]: The skill incorporates AI features that process untrusted data from clinical notes and WhatsApp messages, representing an indirect prompt injection surface.
- Ingestion points: Clinical notes processed in
ClaudeService.summarizeClinicalNotes, patient history inClaudeService.analyzePatientRisk, and incoming WhatsApp text inWhatsAppService.handleIncomingMessage. - Boundary markers: Absent. The code snippets provided do not use specific delimiters or instructions to prevent the AI from following commands embedded within user data.
- Capability inventory: The agent has capabilities to send automated WhatsApp messages, generate tax invoices via AFIP, and create rooms for video consultations.
- Sanitization: The provided services perform standard data handling but do not include explicit sanitization or filtering logic designed to mitigate prompt injection attacks.
Audit Metadata