skill-creator
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts and management tools included in the repository (e.g.,
scripts.aggregate_benchmark,scripts.run_loop) to automate evaluation workflows. - [COMMAND_EXECUTION]: It manages background processes using
nohupto start a local server for the evaluation viewer. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes user-provided evaluation queries and feedback that are then passed to subagents to guide further skill development.
- Ingestion points: Processes content from
feedback.json,eval_set.json, andevals.jsonwithin the skill workspace. - Boundary markers: No specific boundary markers or instructions to ignore embedded commands are defined for ingested feedback.
- Capability inventory: Includes local command execution, subagent orchestration, and file system manipulation across multiple bundled scripts.
- Sanitization: No explicit sanitization or validation steps are described for user-provided feedback before it influences subagent instructions.
Audit Metadata