whatsapp-waha
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: Exposure of hardcoded infrastructure IP address.
- Evidence: The environment variables section in
SKILL.mdlistshttp://172.245.56.50:3000as the server URL for all product integrations, exposing backend infrastructure details. - [PROMPT_INJECTION]: Indirect Prompt Injection vulnerability surface via incoming message processing.
- Ingestion points: Untrusted data enters the context via a WhatsApp webhook at
/api/whatsapp/claude/superseller(SKILL.md). - Boundary markers: Absent; the documentation does not specify the use of delimiters or instructions to ignore embedded commands within processed messages.
- Capability inventory: The agent has access to automated messaging capabilities (
sendText,sendVideo,sendFile) which could be misdirected by injected instructions. - Sanitization: Absent; no validation or sanitization of incoming WhatsApp message content is mentioned before it is processed by the Claude agent.
Audit Metadata