国考岗位辅助

This MCP server provides six tools to help you find and match national civil service positions. It offers multi-criteria filtering, major-based querying, personal condition matching, detailed position info, statistics, and position comparisons. Run locally to power interactive searches and save time when exploring eligible国考岗位.

How to use

You interact with the server through an MCP client to perform practical searches. Use the tools to filter jobs by multiple criteria, search by major keywords, or match positions to your personal profile. You can also view detailed information about a specific position, generate statistics by group, or compare several positions side by side. Focus on actionable tasks like finding all computer science related roles in Beijing, or checking how many positions require a bachelor’s degree in a given机关类别.

Key workflows you’ll perform: 1) filter positions with several conditions such as department, location, education, and major; 2) search for roles by a major keyword; 3) run a one-click personal match using your education, major, political status, location, and work experience; 4) view full details for a chosen position; 5) get aggregated statistics by a chosen grouping field; 6) compare two or more positions to see their differences at a glance.

How to install

pip install -r requirements.txt

# Prepare data file path or environment variable
# Option A: place the data file next to the server and let the program find it by default
# Option B: explicitly set the path via environment variable as shown below
export GUOKAO_EXCEL_PATH=/path/to/2025年国考岗位表.xls

# Run the server locally
python mcp_server.py

# The server exposes an HTTP endpoint and can also run in SSE/stdio mode as configured below.

Additional sections

Environment and data setup

• Ensure the data file 2025年国考岗位表.xls is accessible to the server. It should contain fields such as 部门名称, 招考职位, 工作地点, 学历, 政治面貌, 基层工作最低年限, 专业, 招考人数, etc.

Runtime options

• The server can be run in standard local mode using Python. You can also configure a client to connect via HTTP to a remote MCP endpoint if you deploy it elsewhere. If you use a UI client like Claude Desktop, you can point it to the local SSE endpoint or to a remote URL once configured.

Notes on usage

• Default results are capped at 30 items. You can adjust this with max_results in your queries. Always provide the required fields for each tool to get accurate results.

Security and data considerations

• Keep your data file secure and ensure it only contains permissible, non-sensitive information for your workflow. When running in shared environments, limit access to the server endpoints and use environment variables to configure paths safely.

Available tools

filter_positions

Filters positions by multiple criteria such as organ_type, department, location, education, political_status, work_years, major, exam_type, position_type, min_recruitment, max_recruitment, and max_results to return matching results.

match_by_major

Returns positions that match a given major keyword, with an optional max_results limit.

personal_match

Performs a one-click match for a user based on education, major, political_status, work_years, location, and has_base_project, returning suitable positions.

get_position_detail

Fetches detailed information for a specific position given its department and position_name.

get_statistics

Provides aggregated statistics on positions grouped by a chosen field such as机关类别, 学历, 政治面貌, or 基层工作最低年限.

compare_positions

Compares multiple positions by listing their department and position_name side by side.