openclaw-with-vscode

面向 OpenClaw 的桥接 Skill：通过本地 HTTP 端点，把用户编写、编辑、审查或调试代码的请求转发给 VS Code 中的 Copilot 执行，并返回其回复文本。使用前需安装对应 VS Code 扩展、保持 VS Code 打开，并可用 curl 做健康检查；请求最终由 Copilot 自身服务处理。

diankourenxia

热度 76.08

最近更新 2026-05-18 12:26:23

分类：开发依赖：代码执行 / 第三方 API应用场景：代码生成 / 代码审查 / 调试排错 / API 集成

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add openclaw-with-vscode

Metadata

name: vscode-copilot
description: Bridge between OpenClaw and VS Code Copilot — dispatch coding tasks from any OpenClaw channel to VS Code for execution.
version: 0.2.0
homepage: https://marketplace.visualstudio.com/items?itemName=wodeapp.openclaw-chat
metadata: {"openclaw":{"emoji":"💻","requires":{"bins":["curl"]}}}

VS Code Copilot Bridge

Dispatch coding tasks from OpenClaw to VS Code Copilot for execution. When the user asks you to write, edit, review, or debug code, send the task to VS Code where Copilot will execute it — editing files, running commands, creating code, and more.

将编码任务从 OpenClaw 分发到 VS Code Copilot 执行。当用户要求编写、编辑、审查或调试代码时，将任务发送到 VS Code，由 Copilot 代为执行——编辑文件、运行命令、创建代码等。

Setup / 安装

Install the OpenClaw Chat extension from VS Code Marketplace: 从 VS Code 扩展商店安装 OpenClaw Chat 扩展：

code --install-extension wodeapp.openclaw-chat

Or search "OpenClaw Chat" in VS Code Extensions panel. 或在 VS Code 扩展面板搜索 "OpenClaw Chat" 安装。

The extension starts automatically with VS Code. No manual launch needed. 扩展会随 VS Code 自动启动，无需手动操作。

External Endpoints / 端点

Endpoint	Method	Data Sent / 发送数据
`http://localhost:19836/trigger`	POST	`{"prompt":"..."}` — the coding task / 编码任务
`http://localhost:19836/health`	GET	None / 无

All traffic is local (127.0.0.1). No data leaves the machine. 所有流量均在本地（127.0.0.1），不会发送到外部。

Workflow / 工作流程

Check if the extension is running / 检查扩展是否运行：

curl -s http://localhost:19836/health

If not running, guide the user through Setup above. 如未运行，引导用户完成上方安装步骤。

Send the coding task / 发送编码任务：

curl -s -X POST http://localhost:19836/trigger \
  -H "Content-Type: application/json" \
  -d '{"prompt":"<TASK_DESCRIPTION>"}'

The response JSON contains a response field with Copilot's reply. Display it to the user. 返回的 JSON 包含 response 字段，内含 Copilot 的回复。将其展示给用户。

Example / 示例

User says / 用户说: "帮我写一个防抖函数"

curl -s -X POST http://localhost:19836/trigger \
  -H "Content-Type: application/json" \
  -d '{"prompt":"在当前打开的文件里写一个 TypeScript 防抖函数"}'

Response / 返回：

{"ok":true,"prompt":"...","response":"Here is a debounce function..."}

Security & Privacy / 安全与隐私

All HTTP traffic stays on localhost (127.0.0.1:19836) / 所有流量仅在本地
No data is sent to external servers by this skill / 此 skill 不会向外部发送数据
Copilot processes the request using GitHub's API (standard Copilot behavior) / Copilot 通过 GitHub API 处理请求（标准行为）

Notes / 注意事项

VS Code must be open / VS Code 必须处于打开状态
Copilot Chat should be in Agent mode for full execution / Copilot Chat 应切换到 Agent 模式以获得完整执行能力
Always run health check before dispatching if unsure / 不确定时先执行健康检查
Always show the response to the user / 始终将回复展示给用户