一键导入
contextpilot
Optimize document ordering in LLM context for better retrieval performance
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
Optimize document ordering in LLM context for better retrieval performance
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
| name | contextpilot |
| description | Optimize document ordering in LLM context for better retrieval performance |
| version | 1.0.0 |
| triggers | ["/contextpilot","/cp"] |
You have ContextPilot enabled as a transparent proxy. All your LLM API requests are automatically routed through ContextPilot, which reorders documents in your context window for optimal retrieval performance.
When your request contains multiple documents (in system prompts or tool results), ContextPilot:
<documents>, <files>, etc.), numbered lists, separators, or markdown headers<documents><document>...</document></documents>, <files><file>...</file></files>[1] doc [2] doc [3] doc--- or ===# or ## headersCheck the X-ContextPilot-Result response header:
X-ContextPilot-Result: {"intercepted":true,"documents_reordered":true,"total_documents":5,"sources":{"system":1,"tool_results":1}}
Send these headers with your API requests to control behavior:
| Header | Values | Default |
|---|---|---|
X-ContextPilot-Enabled | true/false | true |
X-ContextPilot-Mode | auto/xml_tag/numbered/separator/markdown_header | auto |
X-ContextPilot-Scope | all/system/tool_results | all |
X-ContextPilot-Alpha | float | 0.001 |