ravi-inbox
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes vendor-specific command-line tools (
ravi inbox sms,ravi inbox email) to retrieve communication data from the Ravi platform. - [DATA_EXFILTRATION]: Accesses sensitive incoming messages, including One-Time Passwords (OTPs), 2FA codes, and account verification links, which are high-value targets for credential harvesting.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection by processing arbitrary text from external SMS and email messages without sanitization.
- Ingestion points: Incoming message bodies retrieved via
ravi inboxinSKILL.md. - Boundary markers: None; message content is processed as raw text without delimiters.
- Capability inventory: Read-only access in this specific skill.
- Sanitization: None; the provided examples use
jqandgrepto extract data directly from untrusted message text.
Audit Metadata