Native codex coordination
Skill Kbediako/evergreen-codex-skills/skills/native-codex-coordination
Evergreen, generally useful Codex skills designed to stay valuable as models improve.
npx -y skills add Kbediako/evergreen-codex-skills --skill native-codex-coordinationAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 21 days oldThe repository was created 21 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Choose the smallest suitable Codex workflow among solo work, bounded native-agent execution, deliberation, native-agent evaluation, skill forward-testing, and waiting. Use when the main uncertainty is which of these coordination shapes fits the task.
SKILL.md
2.1 KB, as published. Nobody here has run it
Native Codex Coordination
Route
Choose one primary workflow:
- Stay solo for small, sequential, tightly coupled, or directly verifiable work.
- Load native-subagents-first for authorized execution or review that splits into independent scopes.
- Load native-agent-deliberation for a consequential decision needing distinct viewpoints and dissent.
- Load native-agent-evals to test or audit the active V2 native-agent behavior.
- Load native-agent-skill-validation to forward-test an existing or changed skill.
- Load wait-for-subagents-patiently for a quiet or over-budget native child.
- Load long-poll-wait for an external job, process, or CI run that needs terminal-state monitoring.
Load any task-domain skill as well. Let the domain skill govern the work and use this skill only to select the coordination shape.
Decide
- Identify the requested outcome and authority boundary.
- Check whether work is independent enough to benefit from native agents.
- Select the narrowest route above.
- Load the selected skill before acting.
- Follow only that skill's workflow and acceptance criteria.
- Verify the integrated result before handoff.
Do not copy lifecycle, model, fanout, budget, brief, or resume rules into this router. Treat native-subagents-first as the sole contract for those concerns.
Boundaries
- Do not trigger generic fanout merely because a task is complex.
- Do not use deliberation as an implementation workflow.
- Do not use evals as production work.
- Do not use native-child waiting rules for external processes.