isaaclab-randomizing-with-events
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The validation section in
SKILL.mdincludes a local execution commanduv run --no-project python tools/skills/cli.py check. This is a standard practice for verifying skill integrity using internal developer tooling. - [INDIRECT_PROMPT_INJECTION]: The
evaluations.mdfile defines scenarios where the agent processes user queries about simulation parameters. The skill lacks explicit boundary markers for these inputs; however, the scope is strictly limited to generating Isaac Lab environment configurations.
Audit Metadata