菜单
Search and extraction strategy. Teaches the researcher which of the 8 web tools to use at each stage of a research cycle.
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
How to orchestrate work across large documents and multi-agent delegations without burning token budgets. Reconnaissance-first, search-then-read, parallel-after-reconnaissance.
Decision criteria for delegating to the writer agent vs synthesising directly in your own response.
When and how to use the document library
How the Coder agent drives an OpenCode session via dispatch_coder — the engineering loop, when to dispatch, and how to verify the result is real (vs. claimed).
Maintain a personal index of interesting things, open questions, and follow-ups across sessions
When and how to render a promoted export to PDF via the export_pdf tool. Reads the writer's draft → export → PDF lifecycle and the operator-facing failure modes when the host is missing pandoc or tectonic.
| name | web-research |
| description | Search and extraction strategy. Teaches the researcher which of the 8 web tools to use at each stage of a research cycle. |
| agents | researcher, cognitive |
You have 8 web tools arranged in tiers. Pick the right tool for the task:
| Tool | Best for | Notes |
|---|---|---|
| brave_answer | Self-contained factual questions ("what is X", "when did Y") | Fastest, cheapest for simple facts |
| brave_llm_context | Default hot path — need raw context to reason over | Machine-optimised search results |
| brave_web_search | Broad discovery with titles, URLs, and snippets | Good for finding multiple sources |
| brave_news_search | Time-sensitive queries, current events | Returns recent news articles |
| brave_summarizer | Need citations + follow-up threads | Search + summary chain |
Factual, self-contained question? → brave_answer
Need raw context to reason over? → brave_llm_context (default)
Broad discovery with snippets? → brave_web_search
Time-sensitive / news? → brave_news_search
Need citations + follow-up threads? → brave_summarizer
Have a URL, need full content? → jina_reader (primary), fetch_url (fallback)
No API keys available? → web_search (DuckDuckGo fallback)
After extraction, check:
Prefer primary sources. When a snippet and a full extraction conflict, trust the full extraction.