israeli-sms-gateway
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data at runtime through the message content (
args.message) and phone numbers (args.to) passed toscripts/send_sms.py. While phone numbers are sanitized using regex inscripts/validate_phone.py, the message body is interpolated directly into API payloads for external gateways. - Ingestion points: Command-line arguments
args.toandargs.messageinscripts/send_sms.py. - Boundary markers: None present to delimit user input from agent instructions.
- Capability inventory: Performs network POST requests to external SMS provider APIs (Twilio, InforUMobile, SMS4Free).
- Sanitization: Phone numbers undergo regex-based normalization to +972 format; message content is not sanitized for prompt injection patterns.
- [COMMAND_EXECUTION]: The skill provides Python scripts intended for execution in a shell environment. These scripts (
scripts/send_sms.pyandscripts/validate_phone.py) perform logic based on user input and environment variables, which is standard for developer tooling. - [EXTERNAL_DOWNLOADS]: The skill uses the
requestslibrary to communicate with external SMS gateway providers, including Twilio, InforUMobile, and SMS4Free. These are well-known or official service providers for the skill's stated purpose, and the communication is documented neutrally as a core functional requirement.
Audit Metadata