The Agent Skills Directory

[SAFE]: No security issues were detected. The skill consists of instructional text for an AI agent to manage multi-model workflows and use execution tools responsibly.
[COMMAND_EXECUTION]: The skill mentions using a codex exec command for model delegation. This is a standard platform capability and the instructions explicitly advise the agent to request necessary permissions from the user rather than bypassing runtime restrictions.
[INDIRECT_PROMPT_INJECTION]: The skill involves routing tasks to sub-agents and processing their outputs. It mitigates potential injection risks by instructing the primary model ('Sol') to inspect all executor results, integrate them, and perform a final review before completion.

orchestrate-models