i4h-workflow-dataset-teleop

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill clones the workflow repository from "https://github.com/isaac-for-healthcare/i4h-workflows". This is an official resource from the developer's ecosystem.- [COMMAND_EXECUTION]: The skill executes bash scripts like "arena/run.sh" and "stop.sh" to control the simulation and recording processes. This is standard functionality for the intended use case.- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes external configuration and log files.
  • Ingestion points: Environment YAML configurations and teleoperation logs in "SKILL.md".
  • Boundary markers: Not used.
  • Capability inventory: Shell command execution, directory operations, and repository cloning.
  • Sanitization: No specific sanitization of external data is documented.
  • Note: This is a low-risk surface in a developer tool where data is typically user-generated or local.- [SAFE]: No malicious patterns, data exfiltration, or unauthorized persistence mechanisms were found. All activities are consistent with the documented purpose of recording robot demonstrations.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 04:01 AM
Security Audit — agent-trust-hub — i4h-workflow-dataset-teleop