DeepCopilot
Copilot-style AI coding agent for VS Code, powered by DeepSeek V4.
Install / Use
npx skills add deep-copilot/DeepCopilotInstalls into whichever agent you are using.
Other
Other agent config
Quality Score
Category
AI & Machine LearningSupported Platforms
Skill content
View source on GitHubDeep Copilot
<p align="center"> <img src="imgs/main_logo.png" alt="Deep Copilot" width="100%"/> </p> <p align="center"> <b>基于 DeepSeek API 的 VS Code 对话式 AI 编程助手扩展</b><br/> <sub>A VS Code extension for conversational AI-assisted coding via the DeepSeek API</sub> </p> <p align="center"> <a href="https://code.visualstudio.com/"><img src="https://img.shields.io/badge/VS%20Code-%E2%89%A51.95.0-blue" alt="VS Code"/></a> <a href="https://github.com/ZhouChaunge/DeepCopilot/releases"><img src="https://img.shields.io/github/v/release/ZhouChaunge/DeepCopilot?label=version&color=success" alt="Version"/></a> <a href="https://github.com/ZhouChaunge/DeepCopilot/stargazers"><img src="https://img.shields.io/github/stars/ZhouChaunge/DeepCopilot?style=flat&color=yellow" alt="GitHub stars"/></a> <a href="https://marketplace.visualstudio.com/items?itemName=ZhouChaunge.deep-copilot"><img src="https://img.shields.io/visual-studio-marketplace/i/ZhouChaunge.deep-copilot?label=installs&color=brightgreen" alt="VS Marketplace Installs"/></a> <a href="./LICENSE"><img src="https://img.shields.io/badge/license-MIT-green" alt="License"/></a> </p>Deep Copilot 是一个 VS Code 扩展,在侧边栏提供基于大语言模型的对话式编程辅助。它通过 DeepSeek API(兼容 OpenAI 协议)与模型交互,支持文件读写、代码搜索、Shell 命令执行等工具调用,结果以流式方式实时渲染。扩展无运行时 npm 依赖,基于 VS Code Extension API 与 Node.js 内置模块实现,无需在宿主机额外部署任何服务。
Deep Copilot is a VS Code extension that provides LLM-based conversational coding assistance through the sidebar. It interacts with models via the DeepSeek API (OpenAI-compatible protocol), supporting tool calls for file read/write, code search, and shell command execution — all streamed in real time. The extension has no runtime npm dependencies and is built on the VS Code Extension API and Node.js built-ins; no additional service deployment is required.
🔑 API Keys Required · 需要配置的 API Key
开始使用前只需准备以下 Key(最少只需第一个):
You only need the following keys to get started — at minimum just the first one:
| # | API Key | 用途 · Purpose | 获取地址 · Get it here | 是否必须 · Required |
|---|---|---|---|---|
| 1 | DeepSeek API Key | 驱动 AI 对话与 Agent 工具调用 · Powers all AI chat & agent tool calls | platform.deepseek.com/api_keys | ✅ 必须 · Required |
| 2 | Tavily API Key | 启用联网搜索工具 web_search · Enables the web_search tool | app.tavily.com | ⚙️ 可选 · Optional |
如何填入 · How to set them
中文:
- 安装扩展后,点击侧边栏活动栏的 🐋 图标打开 Deep Copilot 面板
- 点击面板右下角 🔑 按钮 → 粘贴 DeepSeek API Key → 保存
- 若需要联网搜索,在同一弹窗里继续填入 Tavily API Key
English:
- After installing, click the 🐋 icon in the activity bar to open the Deep Copilot panel
- Click the 🔑 button in the bottom-right of the panel → paste your DeepSeek API Key → save
- For web search, fill in your Tavily API Key in the same dialog
国内用户:若
api.deepseek.com连接不稳定,在 🔑 弹窗中将 Base URL 改为https://api.deepseeki.com。
China users: ifapi.deepseek.comis slow, set Base URL tohttps://api.deepseeki.comin the 🔑 dialog.
📑 Table of Contents · 目录
- API Keys · 需要配置的 Key
- Highlights · 亮点
- Quick Start · 快速开始
- Build from Source · 源码构建
- Configuration · 配置
- Keybindings · 快捷键
- Tools · 工具列表
- Architecture · 架构
- Project Structure · 项目结构
- Development · 开发
- Troubleshooting · 故障排查
- Changelog · 更新日志
- Star History
- License
✨ Highlights · 亮点
| English | 中文 |
| -------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- |
| Agentic loop with multi-turn tool calling on DeepSeek V4 (Pro / Flash / Reasoner) | 与 DeepSeek V4(Pro / Flash / Reasoner)多轮tool-calling 循环 |
| File tools: read, write, str-replace, apply_patch, list dir, find files, ripgrep search | 文件工具:读 / 写 / 精准替换 / apply_patch / 列目录 / 查找 / 全文搜索 |
| Shell tool with configurable approval policy | 终端工具,按审批策略弹窗确认 |
| Web search via Tavily (optional API key) | 网络搜索(Tavily,可选 Key) |
| Plan & Todos panel — agent maintains a structured plan you can watch tick off | Plan & Todos 面板,Agent 自维护结构化任务并实时勾选 |
| Revert last turn — one-click rollback of all file edits in the current agent turn | 一键回滚当前 Agent 轮次对文件的所有修改 |
| Pending edits panel — GH-Copilot-style review popover above the composer: per-file +N/-M, hover Keep/Discard, click row to open native diff editor; survives turn end | 待审编辑面板—输入框上方弹出 GH Copilot 风格的复检面板:逐文件 +N/-M、悬停保留/丢弃、点击行打开原生 Diff 编辑器;对话结束后仍可继续点击 |
| User memory (~/.deepcopilot/memory.md) — cross-project preferences in every system prompt | 用户记忆(~/.deepcopilot/memory.md),跨项目偏好自动注入系统提示词 |
| MCP client — connect any MCP-compatible tool server via deepseekAgent.mcp.servers | MCP 客户端,通过 deepseekAgent.mcp.servers 连接任意 MCP 工具服务器 |
| Post-tool hooks — run scripts after any tool call; output injected into model context | 工具后置钩子,工具调用后自动执行脚本,输出注入模型上下文 |
| Post-edit LSP diagnostics appended to every edit so the model can self-verify | 每次编辑后自动附加LSP 诊断,模型可自行校验 |
| Per-workspace session history with search, rename, delete | 每工作区独立的会话历史,可搜索 / 重命名 / 删除 |
| Parallel sessions — switch away from a running task and start another; live replay on return | 多会话并行:任务跑着可以切走开新对话,回来自动回放进度 |
| Streaming output with reasoning expander, blinking cursor, top progress bar | 流式输出:思维链可展开、闪烁光标、顶部进度条 |
| HTML rendering — model responses render full Markdown + HTML; math via KaTeX | HTML 渲染:响应支持完整 Markdown + HTML + KaTeX 数学公式 |
| Account balance display in footer (click to refresh) | 账户余额实时显示在底栏(点击刷新) |
| Auto-grow input — textarea grows with content, GH Copilot style | 自适应输入框:随内容自动增高,对齐 GH Copilot 体验 |
| Approval modes: Manual / Auto-Edit / Autopilot / Read-Only | 审批模式:手动 / 自动编辑 / 全自动 / 只读 |
| Cost telemetry in CNY shown in the footer | 底栏显示token 数与人民币成本 |
| Slash commands (/explain, /fix, /tests …), @ file refs and # context refs — pick #file, #selection, #editor, #problems, #changes, #terminal, #symbol:Foo, #fetch:URL from the input | 斜杠命令 + @ 文件引用 + # 上下文引用(在输入框键入 # 即可挑选文件 / 选区 / 编辑器 / 诊断 / Git 改动 / 终端 / 符号 / 抓取 URL) |
| Smart code-block actions: Run in terminal · Insert · Copy · Fold long blocks | 代码块操作:在终端运行 / 插入 / 复制 / 长代码折叠 |
| Bilingual UI + locale-aware fonts: auto follows VS Code locale (zh-cn / en) — Chinese font stack for CJK, Latin font stack for EN | 中英双语 UI + 语言感知字体:跟随 VS Code 语言自动切换界面文案与字体(中文环境使用微软雅黑/苹方,英文环境使用 Segoe UI/Inter) |
| Skills system — define reusable SKILL.md packs in ~/.deepcopilot/skills (or ~/.claude/skills, ~/.copilot/skills); YAML frontmatter for workspace gating, invoke via /skill or the model's skill_invoke tool | 技能系统:在 ~/.deepcopilot/skills 等目录放置 SKILL.md 技能包,支持 YAML 元数据与工作区门控,/skill 命令或 skill_invoke 工具均可唤起 |
| Inline FIM completions — DeepSeek ghost-text suggestions as you type; Tab to accept; off by default (deepCopilot.inlineCompletion.enable) | 行内 FIM 补全:基于 DeepSeek 的编辑器幽灵文字建议,Tab 接受;默认关闭,开关:deepCopilot.inlineCompletion.enable |
| Plan mode — read-only investigation mode; agent can read/search but never write or run shell commands | Plan 只读模式:仅允许读文件/搜索,拒绝任何写操作,适合调查代码结构时使用 |
| Ecosystem AI-rule discovery — auto-injects DEEPCOPILOT.md, .github/copilot-instructions.md, AGENTS.md, .cursor/rules/*.mdc, CLAUDE.md | 项目规则自动发现:自动注入工作区下的主流 AI 规则文件(GitHub Copilot / Cursor / Claude 等),让模型了解项目约定 |
| Context window management (new in 0.41.0) — structure-aware truncation, per-file dedup, rolling summary; /compact [focus] force-compacts, /context opens a token breakdown, /fork [name] branches a new session from any message | 上下文窗口管理 (0.41.0 新增):结构感知截断、同文件去重、滚动摘要;/compact [focus] 立即压缩,/context 查看 token 占用细分,/fork [name] 从任一消息派生新会话 |
| Footer context ring (new in 0.41.0) — ring indicator next to the footer ramps green → yellow → orange → red across 60 / 85 / 100% thresholds; click to open the same breakdown as /context | 底部 Context Ring (0.41.0 新增):底栏环形进度指示,按 60% / 85% / 100% 阈值由绿渐变到红;点击展开与 /context 一致的占用细分 |
| Archive = pure export (updated in 0.41.6) — "Archive" now writes a Markdown snapshot under .deep-copilot/archives/ and leaves the session alone; old soft-hidden sessions are auto-unarchived on first launch | 存档 = 纯导出 (0.41.6 调整):点「存档」只生成 .deep-copilot/archives/ 下的 Markdown 快照,不再隐藏会话;老版本被隐藏的会话升级后首次启动自动还原。 |
| Watchdog turns stay on duty (new in 0.41.6) — when the agent monitors a long-running background job (training, build, dev server) via read_terminal, the turn refuses to end until the job finishes or the 4 h per-turn budget elapses | 值守任务不提前退出 (0.41.6 新增):模型用 read_terminal 盯住某个后台任务(训练 / 构建 / 开发服务)后,turn 不会在“我会持续盯”之后静默断开,直到任务结束或命中 4h 本轮预算 |
🚀 Quick Start · 快速开始
Option 1 — VS Code Marketplace · 从扩展商城安装(推荐)
- Open VS Code → Extensions (
Ctrl/Cmd+Shift+X) → Search Deep Copilot → Install. 打开 VS Code → 扩展面板(Ctrl/Cmd+Shift+X)→ 搜索 Deep Copilot → 安装。
Option 2 — Install the prebuilt VSIX · 安装预构建 VSIX
# 从 GitHub Releases 下载最新版 / Download from GitHub Releases:
# https://github.com/ZhouChaunge/DeepCopilot/releases
code --install-extension deep-copilot-0.41.6.vsix
Or in VS Code: Extensions view → ⋯ menu → Install from VSIX... and pick the file.
或在扩展面板右上角 ⋯ → Install from VSIX... 选择文件。
Step 2 — Set the API key · 配置 API Key
- Click the 🐋 Deep Copilot icon in the activity bar to open the chat panel.点击活动栏中的 🐋 Deep Copilot 图标,打开聊天面板。
- Click the 🔑 button at the bottom right of the panel, paste your DeepSeek API key.点击面板右下角 🔑 按钮,粘贴你的 DeepSeek API Key。
- Start chatting! 开始对话!
🛠 Build from Source · 源码构建
Prerequisites · 前置依赖
| Tool | Version | Note | | ----------------- | ------- | ------------------ | | Node.js | ≥ 18 | esbuild + vsce | | npm | ≥ 9 | comes with Node | | VS Code | ≥ 1.95 | extension host | | Git | any | optional, to clone |
Steps · 构建步骤
# 1. Clone the repo · 克隆仓库
git clone https://github.com/ZhouChaunge/DeepCopilot.git
cd DeepCopilot
# 2. Install dependencies · 安装依赖
# (only devDependencies: esbuild + vsce + @types — runtime is pure VS Code API)
# (只有开发依赖;运行时仅用 VS Code API,无运行时 npm 依赖)
npm i
Truncated for display — read the full file on GitHub.
Related Skills
caveman
107.2k🪨 why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by talking like a caveman.
claude-mem
94.4kPersistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
Agent-Reach
84.4kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
Understand-Anything
83.6kGraphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.
