SUNTAO0213 commited on
Commit
276c676
·
1 Parent(s): f008a84

update readme and mcp demo

Browse files
README.md CHANGED
@@ -422,6 +422,30 @@ Here are some tasks you can try with OWL:
422
 
423
  OWL's MCP integration provides a standardized way for AI models to interact with various tools and data sources:
424
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
425
  Try our comprehensive MCP example in `examples/run_mcp.py` to see these capabilities in action!
426
 
427
  ## Available Toolkits
 
422
 
423
  OWL's MCP integration provides a standardized way for AI models to interact with various tools and data sources:
424
 
425
+ Before using MCP, you need to install Node.js first.
426
+ ### **Install Node.js**
427
+ ### Windows
428
+
429
+ Download the official installer: [Node.js](https://nodejs.org/en).
430
+
431
+ Check "Add to PATH" option during installation.
432
+
433
+ ### Linux
434
+ ```bash
435
+ sudo apt update
436
+ sudo apt install nodejs npm -y
437
+ ```
438
+ ### Mac
439
+ ```bash
440
+ brew install node
441
+ ```
442
+
443
+ ### **Install Playwright MCP Service**
444
+ ```bash
445
+ npm install -g @executeautomation/playwright-mcp-server
446
+ npx playwright install-deps
447
+ ```
448
+
449
  Try our comprehensive MCP example in `examples/run_mcp.py` to see these capabilities in action!
450
 
451
  ## Available Toolkits
README_zh.md CHANGED
@@ -410,6 +410,29 @@ OWL 将自动调用与文档相关的工具来处理文件并提取答案。
410
  ## 模型上下文协议(MCP)
411
 
412
  OWL 的 MCP 集成为 AI 模型与各种工具和数据源的交互提供了标准化的方式。
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
413
 
414
  查看我们的综合示例 `examples/run_mcp.py` 来体验这些功能!
415
 
 
410
  ## 模型上下文协议(MCP)
411
 
412
  OWL 的 MCP 集成为 AI 模型与各种工具和数据源的交互提供了标准化的方式。
413
+ 在使用MCP前,需要先安装Node.js。
414
+ ### **安装 Node.js**
415
+ ### Windows
416
+
417
+ 下载官方安装包:[Node.js](https://nodejs.org/zh-cn)。
418
+
419
+ 安装时,勾选 "Add to PATH" 选项。
420
+
421
+ ### Linux
422
+ ```bash
423
+ sudo apt update
424
+ sudo apt install nodejs npm -y
425
+ ```
426
+ ### Mac
427
+ ```bash
428
+ brew install node
429
+ ```
430
+
431
+ ### **安装 MCP 服务**
432
+ ```bash
433
+ npm install -g @executeautomation/playwright-mcp-server
434
+ npx playwright install-deps
435
+ ```
436
 
437
  查看我们的综合示例 `examples/run_mcp.py` 来体验这些功能!
438
 
examples/mcp_servers_config.json CHANGED
@@ -1,13 +1,5 @@
1
  {
2
  "mcpServers": {
3
- "desktop-commander": {
4
- "command": "npx",
5
- "args": [
6
- "-y",
7
- "@wonderwhy-er/desktop-commander",
8
- "--force-file-protocol"
9
- ]
10
- },
11
  "playwright": {
12
  "command": "npx",
13
  "args": ["-y", "@executeautomation/playwright-mcp-server"]
 
1
  {
2
  "mcpServers": {
 
 
 
 
 
 
 
 
3
  "playwright": {
4
  "command": "npx",
5
  "args": ["-y", "@executeautomation/playwright-mcp-server"]
examples/run_mcp.py CHANGED
@@ -22,27 +22,8 @@ Environment Setup:
22
 
23
  2. MCP Server Setup:
24
 
25
- 2.1 MCP Desktop Commander (File System Service):
26
- Prerequisites: Node.js and npm
27
- ```bash
28
- # Install MCP service
29
- npx -y @smithery/cli install @wonderwhy-er/desktop-commander --client claude
30
- npx -y @wonderwhy-er/desktop-commander setup --force-file-protocol
31
-
32
- # Configure in owl/examples/mcp_servers_config.json:
33
- {
34
- "desktop-commander": {
35
- "command": "npx",
36
- "args": [
37
- "-y",
38
- "@wonderwhy-er/desktop-commander",
39
- "--force-file-protocol"
40
- ]
41
- }
42
- }
43
- ```
44
 
45
- 2.2 MCP Playwright Service:
46
  ```bash
47
  # Install MCP service
48
  npm install -g @executeautomation/playwright-mcp-server
@@ -59,7 +40,7 @@ Environment Setup:
59
  }
60
  ```
61
 
62
- 2.3 MCP Fetch Service (Optional - for better retrieval):
63
  ```bash
64
  # Install MCP service
65
  pip install mcp-server-fetch
@@ -168,8 +149,7 @@ async def main():
168
  question = (
169
  "I'd like a academic report about Andrew Ng, including "
170
  "his research direction, published papers (At least 3),"
171
- " institutions, etc.Then organize the report in Markdown "
172
- "format and save it to my desktop"
173
  )
174
 
175
  # Connect to all MCP toolkits
 
22
 
23
  2. MCP Server Setup:
24
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
25
 
26
+ 2.1 MCP Playwright Service:
27
  ```bash
28
  # Install MCP service
29
  npm install -g @executeautomation/playwright-mcp-server
 
40
  }
41
  ```
42
 
43
+ 2.2 MCP Fetch Service (Optional - for better retrieval):
44
  ```bash
45
  # Install MCP service
46
  pip install mcp-server-fetch
 
149
  question = (
150
  "I'd like a academic report about Andrew Ng, including "
151
  "his research direction, published papers (At least 3),"
152
+ " institutions, etc. "
 
153
  )
154
 
155
  # Connect to all MCP toolkits