Skip to main content
Glama

@mcpx-digital/json-tools

MCP server for everyday JSON/JSONC work on local files and strings.

Validate JSON or JSONC, get/set values with jq-like paths, pretty-print or minify, validate against JSON Schema (Ajv), and diff two documents — from Cursor or any MCP client.

Local only. Tools accept file paths or inline strings. They do not fetch URLs, scrape the web, or call external APIs.

Install (MCPX / npx)

npx -y @mcpx-digital/json-tools

Prepared for npm as @mcpx-digital/json-tools. Until published, run from a local clone.

Related MCP server: Universal JSON Agent MCP

Cursor mcp.json example

{
  "mcpServers": {
    "json-tools": {
      "command": "npx",
      "args": ["-y", "@mcpx-digital/json-tools"]
    }
  }
}

Local clone:

{
  "mcpServers": {
    "json-tools": {
      "command": "node",
      "args": ["/absolute/path/to/json-tools-mcp/index.js"]
    }
  }
}

Tools

Tool

What it does

validate_json

Strict JSON parse/validate

validate_jsonc

JSONC (comments + trailing commas)

json_get

Get value at .foo.bar[0]-style path

json_set

Set value at path (optional write)

pretty_json

Pretty-print

minify_json

Minify

schema_validate

Ajv JSON Schema validation

diff_json

Line-oriented diff of two JSON docs

Example prompts

  • “Validate this JSONC config and pretty-print it”

  • “Get .dependencies.react from package.json”

  • “Diff these two API response fixtures”

  • “Validate data.json against schema.json”

Development

git clone https://github.com/TheoryofShadows/json-tools-mcp.git
cd json-tools-mcp
npm install
npm test
node index.js

Sell / list on MCPX

Suggested listing price: $7.

Install command: npx -y @mcpx-digital/json-tools

License

MIT © TheoryofShadows

Available Tools

8 tools
diff_jsonB

Diff two JSON documents (canonical key-sorted stringify, line-oriented). Local files/strings only. Does not fetch URLs or contact external services.

Inputs: jsonc, leftText, maxDiffs, rightText, +2 more.

json_getA

Get a value at a jq-like path (e.g. .foo.bar[0]). Local files/strings only. Does not fetch URLs or contact external services.

Inputs: path (required), text, filePath.

json_setA

Set a value at a jq-like path; returns updated JSON (optionally write file). Local files/strings only. Does not fetch URLs or contact external services.

Inputs: path (required), text, value, indent, +4 more.

minify_jsonA

Minify JSON to one line. Local files/strings only. Does not fetch URLs or contact external services.

Inputs: text, filePath, outputPath.

pretty_jsonB

Pretty-print JSON with indentation. Local files/strings only. Does not fetch URLs or contact external services.

Inputs: text, indent, filePath, outputPath.

schema_validateA

Validate JSON data against a JSON Schema using Ajv. Local files/strings only. Does not fetch URLs or contact external services.

Inputs: jsonc, dataText, schemaText, dataFilePath, +1 more.

validate_jsonA

Parse and validate strict JSON. Local files/strings only. Does not fetch URLs or contact external services.

Inputs: text, filePath.

validate_jsoncA

Parse JSONC (comments + trailing commas) and validate. Local files/strings only. Does not fetch URLs or contact external services.

Inputs: text, filePath.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.9/5.0

    Scored across 8 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: validation (strict vs JSONC), get/set by path, formatting (pretty/minify), schema validation, and diffing. No two tools overlap in functionality.

    Naming Consistency4/5

    Most tools follow a verb_noun pattern (validate_json, pretty_json, minify_json, diff_json), but json_get and json_set reverse the order to noun_verb, creating a minor inconsistency. Overall the pattern is readable and predictable.

    Tool Count5/5

    Eight tools cover a focused domain of JSON utilities without redundancy. This count is well-scoped for the server's purpose.

    Completeness5/5

    The tool surface covers the core lifecycle of JSON manipulation: validation, access, modification, formatting, schema checking, and comparison. No obvious gaps for the stated domain.

    Maintenance

    ActivitySlowing
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      C
      maintenance
      Enables efficient JSON file editing with targeted read, write, delete, and deep merge operations using dot notation paths, optimized for managing multilingual projects and large configuration files.
      4
      30 npm
      1
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Provides tools to validate, format, and query JSON strings via MCP. Enables AI agents to work with JSON data without needing keys or internet.
      131 npm
      MIT