github / github/copilot-cli

Share sessions across machines, and developers

未關閉
#3,537 1 則留言 2 個 reaction 已指派 0 人 在 GitHub 檢視
area:agents area:enterprise area:sessions
主要語言
Shell
星號
11.2k
分支
1.9k
平均合併
14 小時 16 分鐘
30 天內合併 PR
6

描述

### Describe the feature or problem you'd like to solve

I might work on different machines to solve the same problem. I also have teams working together to solve the same problem. It would be good if discoveries, context, history and failures were shareable across sessions and people.

### Proposed solution

Develop some sort of knowledge pool concept. A "problem space" could be created. This could contain references to multiple git repos, documentation, copilot instructions, subagents, MCP server configs etc etc. If the problem space or knowledge pool was built up over time, it would reduce the effect or feeling that every conversation starts from scratch. "/session" is a partial solution to this on a single developers machine, but it does not seem to scale very well. In terms of cost and resource utilisation, and maybe even model training, building up knowledge and known to be true facts could save lots of compute time and improve UX and satisfaction of development teams. This sort of thing would take copilot from single dev's sidekick or swiss army knife, into a Enterprise productivity solution.

### Example prompts or workflows

Workflow: Developer A defines a subagent, everyone in the team can leverage it.
Workflow: A dev team can work collaboratively on a whole platform, over many git repos. Shared context or even a small model could be generated and maintained that improves and evolves as the application being worked upon evolves.
Workflow: The subagents or personas we create could be orchestrated as a team/group to solve larger problems, covering specific angles, such as QA, SRE, DevOps, Security, etc etc.

### Additional context

_No response_

貢獻指南

開啟貢獻指南

研究方向

未指定實作檔案或測試。首先檢查現有的 /session 行為,以及它如何在本機儲存工作階段內容,接著釐清跨機器與開發者共用問題空間的範圍。完成的要求是:針對共用內容、儲存庫、指示、子代理程式與 MCP 設定,達成一致的設計並訂定驗收標準。

由索引模型根據 Issue 內容生成。

評估

技術堆疊
git, shell
領域
ai, cli
Issue 類型
功能
難度
5/5
預估耗時
一週以上
活躍度
活躍
描述清晰度
需要釐清
新手友好度
25/100

把新 issue 寄到你的電子郵件信箱

精選適合新手參與的 GitHub issue 摘要。