القائمة
Offload verbose research (web, code, files, GitHub, video) to a subagent instead of burning main context window.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Use when you have a written implementation plan to execute in a separate session with review checkpoints
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
Use when executing implementation plans with independent tasks in the current session
Use when you have a spec or requirements for a multi-step task, before touching code
Use when creating new skills, editing existing skills, or verifying skills work before deployment
| name | subagent-research |
| description | Offload verbose research (web, code, files, GitHub, video) to a subagent instead of burning main context window. |
Golden rule: if the answer could be longer than 3 lines, use a subagent.
default subagent with questions + brevity instructionAnswer these questions as concisely as possible (one sentence per question):
1. [precise question]
2. [precise question]
No summarisation needed. Use [tool].
grep inline| Type | Tool | Example question |
|---|---|---|
| Web / general | web_search | "What are the pros/cons of X vs Y?" |
| Code / library internals | code_search | "What's the signature of React.useCallback?" |
| GitHub repos | fetch_content | "What does the README say about installation?" |
| YouTube / video | fetch_content + prompt | Pass user's exact question as prompt |
| File / workspace | bash (grep/find), read | "Where is DB configured in this project?" |
| Complex / multi-tool | combination | "Look up lib docs + check usage in codebase" |
web_search for code questions misses signaturesweb_search on main agent → do it in a subagent instead