Searching protocol for "prompt-execution"
Execute forged prompts exactly as written.
Delegate prompts to sub-tasks
Run tasks via Skill Agent.
Safely commit doc changes after review.
Safely disable model invocation for AI tasks.
Delegate coding tasks to an autonomous Codex agent
Autonomous Claude Code execution from Codex CLI.
Load and execute prompts.
Execute prompts with AI models, in worktrees.
Run one-off AI prompts instantly.
Execute delegated tasks with precision.
Test skill injection in realistic scenarios.