isaaclab-training-rl-agents
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the execution of local Isaac Lab scripts and tools using
uv run. This includes training scripts (isaaclab train), playback (isaaclab play), and maintenance tools (tools/skills/cli.py). These are standard operations within the Isaac Lab environment. - [INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it interpolates user-supplied task names and configuration parameters into shell commands.
- Ingestion points: User-provided inputs for task names and RL library selection as described in
SKILL.mdandevaluations.md. - Boundary markers: None identified; the skill does not instruct the agent to use specific delimiters or safety wrappers for user input.
- Capability inventory: Execution of shell commands via the
uvtool and Python argument list processing. - Sanitization: None identified; the instructions rely on the user providing valid registered task names (e.g.,
Isaac-Cartpole).
Audit Metadata