writing-subagents
Use when creating new subagents, editing existing subagents, or verifying subagents work before deployment
- License 许可
- Apache-2.0
- 最近更新
- 2026-08-20
- Stars
- ★ 66
- Forks
- 9
01 前言 Front Matter
- name
- writing-subagents
- description
- Use when creating new subagents, editing existing subagents, or verifying subagents work before deployment
03 SKILL.md 正文
Writing Subagents
Directory Structure
Put subagent files in guidance/agents/codex/. We have linked .codex/agents/ to this folder, so they can be loaded automatically.
guidance/
agents/
codex/
subagent-name.toml
Subagents File Format
Write TOML files with at least:
name = "subagent-name"
description = "Use when [specific triggering conditions and expected output]"
developer_instructions = '''
# Subagent title
Task-specific instructions here.
'''
Reference docs if deeper format details are needed:
- Codex subagents:
https://developers.openai.com/codex/subagents
04 同仓库更多 Skills
writing-skills
Use when creating new skills, editing existing skills, or verifying skills work before deployment
theorem-search
Use TheoremSearch to find, verify, and cite mathematical theorem statements, named results, proof routes, and literature dependencies.
supervise-run
Use when monitoring a running Eve experiment and recovering from problems while it runs
run-math-proof-quickstart
Use when running a Math Proof quickstart from a user-provided problem: create a local task copy, launch a real EvE attempt, supervise it, and report proof/evaluation results.
run-circle-packing-smoke
Use when a fast end-to-end smoke run is needed for the Eve loop development, will run on circle packing task.
resume-run
Use when continuing an interrupted or paused Eve run from where it stopped (same run, same id), optionally after editing the experiment