jcr_mcp

# JCR Partition Table MCP Server **English** | [English](README.en.md) [](https://opensource.org/licenses/MIT) A Model Context Protocol (MCP) server based on the ShowJCR repository data, providing the latest journal partition table query functionality for large language models.  ## Features ### 🔧 Tools - **search_journal** - Search journal information, including impact factor, partition, warning status, etc. - **get_partition_trends** - Get journal partition change trend analysis - **check_warning_journals** - Query international journal warning list - **compare_journals** - Compare comprehensive information of multiple journals ### 📋 Resources - **jcr://database-info** - Database basic information and statistics ### 💡 Prompts - **journal_analysis_prompt** - Journal analysis dedicated prompt word template ## Data Sources This project is based on the data from the [ShowJCR](https://github.com/hitfyd/ShowJCR) repository, including: - **New Journal Partition Table** (2026 version, 22299 journals + 15 important computer field conferences; warning information is embedded in "Warning Mark: Under Review" since this version) - **CAS Partition Table Upgrade Version** (2025, 2023, 2022) - **JCR Journal Impact Factor** (2024, 2023, 2022) - **International Journal Warning List** (2025, 2024, 2023, 2021, 2020; upstream no longer releases separately since 2026) - **CCF Recommended International Academic Conference and Journal Directory** (2026, 2022) - **High-Quality Science and Technology Journal Classification Directory in Computing Field** (2025, 2022) > 2026 version description: The `check_warning_journals` tool will scan both traditional `GJQKYJMD*` warning list and `XR2026.Warning Mark` field, covering new and old warning sources. ## Installation and Deployment ### 1. Environment Requirements - Python **3.10+** (`mcp` SDK hard requirement) - SQLite3 (usually comes with Python) ### 2. Install Dependencies ```bash pip install -r requirements.txt ``` ### 3. Data Synchronization Synchronize data before running for the first time: ```bash python data_sync.py ``` Choose "1" to synchronize all data and wait for download and import to complete. ### 4. Start Server ```bash python jcr_mcp_server.py ``` ## Client Testing ### Independent Testing ```bash python test_client.py ``` Choose mode: - Mode 1: Automatically test all functions - Mode 2: Interactive query mode ### Claude Desktop Integration Add the following to the Claude Desktop configuration file (`%APPDATA%\Claude\claude_desktop_config.json` or macOS's `~/Library/Application Support/Claude/claude_desktop_config.json`): ```json { "mcpServers": { "jcr-partition": { "command": "/path/to/python", "args": ["/path/to/jcr_mcp_server.py"], "cwd": "/path/to/project" } } } ``` The `claude_desktop_config.json` in the repository can be used as a reference template. ### Claude Code Integration Register using CLI (scope `user` represents global availability): ```bash claude mcp add -s user jcr-partition -- /path/to/python /path/to/jcr_mcp_server.py ``` After registration, confirm the status using `claude mcp list`, which should display `✓ Connected`. ## Usage Examples ### 1. Journal Search ```python # Search for Nature journal result = await session.call_tool("search_journal", { "journal_name": "Nature" }) ``` ### 2. Partition Trend Analysis ```python # Get Science journal partition change trend result = await session.call_tool("get_partition_trends", { "journal_name": "Science" }) ``` ### 3. Journal Comparison ```python # Compare three top journals result = await session.call_tool("compare_journals", { "journal_list": "Nature,Science,Cell" }) ``` ### 4. Warning Journal Query ```python # Query warning journals result = await session.call_tool("check_warning_journals", { "keywords": "MDPI" }) ``` ## Output Examples ### Journal Search Result ``` 📚 Journal Name: NATURE 【2024】（JCR） 📊 Impact Factor: 48.5 🏆 Partition: Q1 📖 Subject Category: MULTIDISCIPLINARY SCIENCES(SCIE) 【2025】（CAS Upgrade Version） 🏆 Partition: 1 [1/118]（Top） 📖 Subject Category: Comprehensive Journals 【2026】（New Journal Partition Table） 🏆 Partition: 1 Zone（Top） 📖 Subject Category: Comprehensive Journals ``` ### Journal Comparison Result ``` 📊 Journal Comparison Analysis Result Journal Name Latest Impact Factor Latest Partition Warning Status ---------------------------------------- Nature 64.8 Q1 Normal Science 56.9 Q1 Normal Cell 64.5 Q1 Normal 💡 Submission Suggestions: ⭐ Nature: Top journal, highly recommended ⭐ Science: Top journal, highly recommended ⭐ Cell: Top journal, highly recommended ``` ## Technical Architecture ### Data Layer - SQLite database stores all partition table data - Supports multiple years of historical data - Automatic data synchronization and verification mechanism ### Service Layer - FastMCP framework builds MCP server - Asynchronous processing improves performance - Complete error handling and logging ### Interface Layer - Standard MCP protocol interface - Supports tools, resources, and prompts - Compatible with various MCP clients ## Extension Instructions ### Add New Data Source 1. Add new data source in `data_sources` dictionary in `data_sync.py` 2. Run data synchronization to update database 3. Update parsing logic in `jcr_mcp_server.py` ### Add New Tool 1. Use `@app.tool()` decorator in `jcr_mcp_server.py` 2. Implement specific query logic 3. Add suitable docstring ### Deploy to Cloud Deploy server to cloud platform, support HTTP transport: ```python app.run(transport="streamable-http", host="0.0.0.0", port=8080) ``` ## Related Links - [ShowJCR Original Project](https://github.com/hitfyd/ShowJCR) - [MCP Official Documentation](https://modelcontextprotocol.io/) - [Claude Desktop MCP Integration Guide](https://claude.ai/docs/mcp) ## License This project is open-sourced under the MIT license. ## Contribution Welcome to submit issues and pull requests to improve this project!