routeros-mobile-networking

Pass

Audited by Gen Agent Trust Hub on Oct 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill documents the built-in RouterOS SMS remote command execution feature, which provides an interface for triggering local scripts through incoming messages.
  • Ingestion Point: Incoming SMS messages processed by the /tool/sms service (documented in references/guides/mobile-networking/sms.md).
  • Boundary Markers: The protocol requires a shared secret password to be included in the message body for validation.
  • Capability Inventory: Authorized messages can execute any script available in the router's /system/script repository, which may include file operations or network commands.
  • Sanitization: Content is decoded using the GSM7 alphabet; the documentation explicitly warns users that message content is not authenticated beyond the shared secret.
  • [DATA_EXFILTRATION]: The skill includes multiple tracking examples that transmit device location data (latitude, longitude, altitude, and speed) to external endpoints.
  • Network Operations: Utilizes the /tool/fetch command for HTTP POST requests and /iot/mqtt/publish for MQTT telemetry (e.g., in references/guides/mobile-networking/gps/gps-tracking-using-http-post.md).
  • Targets: All examples use generic placeholder addresses such as YOURSERVER.com and change-this-URL-accordingly.
  • Context: This data transmission is the primary and intended purpose of the tracking guides, representing standard IoT functionality for the target platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 18, 2026, 04:40 PM