json -filter

This MCP server provides JSON schema generation and filtering capabilities for local files and remote HTTP/HTTPS endpoints. It exposes three primary tools: json_schema, json_filter, and json_dry_run. Use json_schema to generate TypeScript interfaces from a JSON payload, json_filter to extract specific fields from large JSON datasets (with optional chunking), and json_dry_run to estimate data size and determine sensible chunking before filtering. The server supports both local and remote data sources, automatic chunking for large responses, and a 50 MB per-source size limit with safeguards and clear error messages to help diagnose issues. Integrations with Claude Desktop and Claude Code are supported for seamless MCP workflows.

Prerequisites:

• Node.js and npm installed on your system (recommended recent LTS version).
• Basic familiarity with running npm or npx commands.

Installation steps:

1. Ensure Node.js is installed. Verify with: node -v npm -v

2. Install or run via NPX (no local install required): npx json-mcp-filter@latest

3. If you prefer a local install from source (optional): git clone <repository-url> cd json-mcp-filter npm install npm run build npm run start

4. For integration in Claude Desktop/Code, use the provided MCP config snippet to register the server. The server can be run using the npx-based command above, which fetches the latest release on each use.

Tips and caveats:

• The tool enforces a 50 MB maximum size per data source. If your input is larger, use json_dry_run to estimate chunking and then json_filter with chunkIndex to process in smaller pieces.
• When working with large remote datasets, prefer json_dry_run first to understand chunk recommendations before performing actual filtering.
• Remote URLs are fetched over HTTP/HTTPS; ensure URLs are trusted and compliant with API terms of service.
• The json_schema tool generates TypeScript interfaces from JSON data for type safety, while json_filter focuses on shape-based field extraction across nested structures and arrays.
• If you encounter parsing or shape-detection errors, check for mixed content types or invalid JSON payloads in the source data.