一键导入
finish
Complete the task with a final answer and save reusable subagents as skills. Use this when you have successfully solved the question.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Complete the task with a final answer and save reusable subagents as skills. Use this when you have successfully solved the question.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Create a new Python subagent file in the workspace. After creating, use run_subagent to test it (NOT use_saved_subagent).
List all previously saved subagent skills. Use this to check if a suitable subagent already exists before creating a new one.
Edit an existing subagent file by replacing specific code snippets. Supports both workspace files and saved skills.
Run a subagent - either a newly created one from workspace or a saved subagent from previous sessions.
View the source code of a saved subagent skill. Use this to understand how an existing subagent works before reusing or adapting it.
**Problem Category**: Audio file transcription using OpenAI Whisper
| name | finish |
| description | Complete the task with a final answer and save reusable subagents as skills. Use this when you have successfully solved the question. |
Complete the task with a final answer and save reusable subagents as skills.
Both answer and subagents are required keys. Omitting either will cause a validation error.
| Parameter | Type | Description |
|---|---|---|
| answer | str | A brief final answer (1-3 sentences). Do NOT put full reports or large content here. |
| subagents | list | List of subagents to save, or [] if none worth saving. |
Do NOT mix skill_name and supersedes in the same entry — they are mutually exclusive.
You created multiple new subagents with create_subagent and want to save them all.
<action>finish</action>
<params>
{"answer": "...", "subagents": [
{"entry_file": "researcher.py", "description": "..."},
{"entry_file": "analyzer.py", "description": "..."},
{"entry_file": "visualizer.py", "description": "..."}
]}
</params>
modify_subagent)You used modify_subagent with skill_name to fix saved skills. Include skill_name so changes are saved back to the original skill directories.
<action>finish</action>
<params>
{"answer": "...", "subagents": [
{"entry_file": "researcher.py", "description": "...", "skill_name": "researcher"},
{"entry_file": "analyzer.py", "description": "...", "skill_name": "analyzer"}
]}
</params>
create_subagent + supersedes)You created new subagents to replace old broken skills. The old skills will be deleted.
<action>finish</action>
<params>
{"answer": "...", "subagents": [
{"entry_file": "researcher.py", "description": "...\n**Improved From**: ...", "supersedes": "old_researcher"},
{"entry_file": "analyzer.py", "description": "..."}
]}
</params>
Use this when: (a) existing saved subagents already solved the problem without any new code or modifications, or (b) the subagents completely failed and contain no useful logic. In all other cases, you MUST save ALL reusable subagents — review every subagent you created in this session and save each one that has value. See Important Notes.
<action>finish</action>
<params>
{"answer": "...", "subagents": []}
</params>
| Field | Type | Required | Description |
|---|---|---|---|
| entry_file | str | Yes | The exact .py filename from create_subagent. Must match a file in the workspace. |
| description | str | Yes | What problems this subagent solves. If supersedes is set, MUST include **Improved From**. |
| skill_name | str | No | Only for Pattern 2. The saved skill name you modified via modify_subagent. Cannot be used with supersedes. |
| supersedes | str | No | Only for Pattern 3. The old skill name to delete and replace. Cannot be used with skill_name. |
**Problem Category**: ...
**Applicable Questions**: ...
**Key Features**: ...
**Skills Used**: ...
**Reasoning Pattern**: ...
**Input Format**: ...
**Output Format**: ...
**Improved From**: [Only if supersedes is set. Explain what the old skill got wrong and what this version fixes.]
answer must be brief (1-3 sentences). Large content causes JSON parsing failures.entry_file must be the exact .py filename you used in create_subagent.skill_name and supersedes are mutually exclusive: use one or the other, never both.finish, review EVERY subagent you created in this session (check all .py files in the workspace). Save each one that has value — do not save only the main one. Even helper subagents or narrowly-focused subagents may be useful for future tasks.view_subagent_code to study its structure, API calls, and workarounds, then create_subagent to build a new one based on it..py files are copied into each saved skill directory.