The Agent Skills Directory

[SAFE]: The skill serves as a logic-based orchestrator for other skills. It processes input data to determine skill activation order, dependency resolution, and risk summaries.
[COMMAND_EXECUTION]: The skill requires the bash tool as defined in SKILL.md to execute its run.py script. The script uses standard Python libraries for JSON processing and logic, with no evidence of arbitrary or dangerous command execution.
[DATA_EXFILTRATION]: No network operations or unauthorized file access patterns were found. The script interacts only with standard input/output and local environment variables via a helper utility.
[INDIRECT_PROMPT_INJECTION]: The skill processes skill metadata which could theoretically contain instructions. However, the logic is restricted to non-executable fields (names, scores, dependency lists) and the output is a structured plan, which significantly limits the surface for indirect injection attacks.

meta-orchestrator