edge-computing

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill facilitates the deployment and execution of custom code on remote edge devices via the OpenClaw CLI and API.
  • Evidence: Mentions 'edge-compute deploy' and 'client.deploy_function' for running scripts on IoT devices.
  • [PROMPT_INJECTION]: The skill processes untrusted real-time data from IoT sensors, creating a surface for indirect prompt injection.
  • Ingestion points: Data streams from sensors and cameras as described in SKILL.md.
  • Boundary markers: No delimiters or instructions to ignore embedded commands are specified for the external data processing.
  • Capability inventory: Includes remote function deployment, status polling, and metrics retrieval.
  • Sanitization: No explicit validation or sanitization of input data is documented in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 10:46 PM