Zoho Timeline MCP Server
Enables Claude Desktop to interactively capture and analyze Zoho CRM timeline screenshots, automatically extracting timeline events, workflows, and functions through OCR and vision API, with export to JSON and Markdown formats.
README
Zoho Timeline MCP Server
An MCP (Model Context Protocol) server that enables Claude Desktop to interactively capture and analyze Zoho CRM timeline screenshots.
Features
- Interactive Screenshot Capture: Click on Zoho CRM windows and let Claude capture them
- Automatic OCR & Parsing: Extracts timeline events, workflows, and functions
- Chronological Analysis: Orders events and identifies patterns
- Export to Multiple Formats: JSON and Markdown reports
Prerequisites
- Python 3.10+
- Claude Desktop (with MCP support)
- Anthropic API Key
- Tesseract OCR (for text extraction)
- Windows: Download from https://github.com/UB-Mannheim/tesseract/wiki
- Add to PATH:
C:\Program Files\Tesseract-OCR
Installation
1. Install Tesseract OCR
Download and install Tesseract from: https://github.com/UB-Mannheim/tesseract/wiki
Make sure to add it to your PATH during installation.
2. Clone or Download This Project
cd %USERPROFILE%\Documents
# If you received this as a zip, extract it here
# Or if in a git repo: git clone <repo-url> zoho-timeline-mcp
cd zoho-timeline-mcp
3. Create Virtual Environment
python -m venv venv
venv\Scripts\activate
4. Install Dependencies
pip install -e .
5. Install Windows-Specific Dependencies
For active window capture on Windows:
pip install pywin32
6. Set Up Anthropic API Key
Create a .env file in the project root:
ANTHROPIC_API_KEY=your_api_key_here
Or set it as an environment variable:
setx ANTHROPIC_API_KEY "your_api_key_here"
Configuration for Claude Desktop
Add this to your Claude Desktop configuration file:
Location: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"zoho-timeline": {
"command": "python",
"args": [
"-m",
"zoho_timeline_mcp.server"
],
"cwd": "C:\\Users\\YourUsername\\Documents\\zoho-timeline-mcp",
"env": {
"ANTHROPIC_API_KEY": "your_api_key_here"
}
}
}
}
Important: Replace YourUsername and your_api_key_here with your actual values.
Usage
1. Start Claude Desktop
After configuring, restart Claude Desktop. The MCP server will start automatically.
2. Capture Screenshots
In Claude Desktop, say:
Capture a screenshot of the Zoho timeline
Claude will:
- Capture the active window
- Save it temporarily
- Ask if you want to capture more
3. Capture Multiple Screenshots
Keep capturing:
Capture another screenshot
Or click on the next part of the timeline and say:
Capture this screen too
4. Process Timeline
When done capturing:
Process the timeline screenshots
Claude will:
- Analyze all screenshots using vision API
- Extract timeline events chronologically
- Identify workflows and functions
- Return comprehensive JSON and report
5. Export Results
Export the results to files
Files are saved to: %USERPROFILE%\Documents\ZohoTimelineAnalysis\
Available Tools
The MCP server provides these tools to Claude:
capture_screenshot
Captures a screenshot of the active window or full screen.
Parameters:
capture_type: "active_window" or "full_screen" (default: "active_window")
list_screenshots
Lists all captured screenshots with IDs and timestamps.
process_timeline
Processes all captured screenshots and extracts timeline data.
Parameters:
api_key: Optional Anthropic API key (uses env var if not provided)
clear_screenshots
Clears all captured screenshots to start fresh.
export_results
Exports timeline analysis to JSON and Markdown files.
Parameters:
output_dir: Optional output directory (default: Documents/ZohoTimelineAnalysis)
Output Format
Timeline JSON
{
"lead_timeline": {
"date": "Nov 25, 2025",
"events": [
{
"timestamp": "02:30 PM",
"event_type": "Lead Created",
"action": "Lead Created",
"triggered_by": "Function",
"function_name": "Leads CR Process Power Almanac Files"
},
...
]
}
}
Workflows & Functions JSON
{
"workflow_rules": {
"count": 13,
"list": [
"Leads - Cr - Format Phone",
...
]
},
"functions": {
"count": 9,
"list": [
"Lead_VerifyEmailHunterIO",
...
]
}
}
Troubleshooting
Screenshots Not Capturing
- Make sure the Zoho window is active and visible
- Try full screen mode:
capture_type: "full_screen" - Check that mss library is installed:
pip install mss
pywin32 Errors
If you get errors about pywin32:
python venv\Scripts\pywin32_postinstall.py -install
API Key Issues
Verify your API key is set:
echo %ANTHROPIC_API_KEY%
Tesseract Not Found
Make sure Tesseract is installed and in your PATH:
tesseract --version
If not found, add to PATH:
C:\Program Files\Tesseract-OCR
MCP Server Not Starting
- Check Claude Desktop logs:
%APPDATA%\Claude\logs - Test the server manually:
python -m zoho_timeline_mcp.server - Verify configuration file syntax (must be valid JSON)
Example Workflow
User: "Capture the current Zoho timeline window"
Claude: [captures screenshot] ✓ Screenshot 1 captured. Ready for another?
User: "Scroll down and capture the next section"
Claude: [captures screenshot] ✓ Screenshot 2 captured. Continue?
User: "One more"
Claude: [captures screenshot] ✓ Screenshot 3 captured. Ready to process?
User: "Process all the screenshots"
Claude: [processes with vision API] ✓ Timeline analyzed!
Found 13 workflow rules and 9 functions...
[displays full report]
User: "Export these results"
Claude: ✓ Saved to Documents\ZohoTimelineAnalysis\
Development
Running Tests
pytest
Code Formatting
black src/
License
MIT License - feel free to modify and use as needed.
Support
For issues or questions:
- Check the troubleshooting section
- Review Claude Desktop MCP logs
- Test the server in standalone mode
Roadmap
- [ ] Support for multiple monitors
- [ ] Automatic scroll detection
- [ ] Batch processing of saved screenshots
- [ ] Export to Excel format
- [ ] Template-based report generation
- [ ] Integration with Zoho API for validation
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。