Install
$ agentstack add mcp-siddiqui-zeeshan-datagrip-database-mcp ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ● Network access Used
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Database MCP Tools
A DataGrip plugin that extends its built-in MCP server with 6 database tools. AI assistants like Claude Code can query databases through DataGrip's existing connections -- SSH tunnels, SSL, and auth included -- with zero extra config.
Requirements
- DataGrip 2025.3+
- JDK 17+
Building
./gradlew buildPlugin
The plugin zip is written to build/distributions/.
Installation
- In DataGrip, go to Settings > Plugins.
- Click the gear icon and select Install Plugin from Disk....
- Select the zip from
build/distributions/. - Restart DataGrip.
Setup
- Enable the MCP server in DataGrip: Settings > Tools > MCP Server (requires DataGrip 2025.3+).
- Datasources auto-connect on demand -- no manual connect step needed (passwords must be saved in DataGrip).
- Configure Claude Code (or another MCP client) to use DataGrip's MCP server.
Option A: Database tools only (recommended)
Use the filter proxy to expose only the 6 database tools. This strips DataGrip's built-in IDE tools from the MCP response, keeping your AI assistant's context clean.
Download the proxy script:
curl -o mcp-filter-proxy.py https://raw.githubusercontent.com/siddiqui-zeeshan/datagrip-database-mcp/main/mcp-filter-proxy.py
{
"mcpServers": {
"datagrip-db": {
"command": "python3",
"args": [
"/path/to/mcp-filter-proxy.py",
"/path/to/java",
"-classpath",
"/path/to/mcpserver-classpath",
"com.intellij.mcpserver.stdio.McpStdioRunnerKt"
],
"env": {
"IJ_MCP_SERVER_PORT": ""
}
}
}
}
See [Classpath and port](#classpath-and-port) below for how to find the correct values.
Option B: All tools (database + built-in IDE tools)
Connect directly to DataGrip's MCP server without filtering. This exposes all built-in IDE tools alongside the database tools.
{
"mcpServers": {
"datagrip": {
"command": "/path/to/java",
"args": [
"-classpath",
"/path/to/mcpserver-classpath",
"com.intellij.mcpserver.stdio.McpStdioRunnerKt"
],
"env": {
"IJ_MCP_SERVER_PORT": ""
}
}
}
}
Classpath and port
The MCP server runs via DataGrip's bundled JRE, not the DataGrip CLI. You need two things:
- Java binary:
/Contents/jbr/Contents/Home/bin/java(macOS) or the equivalent on your OS. - Classpath: the JARs under
/Contents/plugins/mcpserver/lib/and/Contents/lib/. See the [global MCP setup](#classpath-and-port) section for details. - Port: set
IJ_MCP_SERVER_PORTto the port shown in DataGrip under Settings > Tools > MCP Server.
| OS | DataGrip install path | |----|------| | macOS (Toolbox) | ~/Library/Application Support/JetBrains/Toolbox/apps/datagrip/ | | macOS (standalone) | /Applications/DataGrip.app/ | | Linux (Toolbox) | ~/.local/share/JetBrains/Toolbox/apps/datagrip/ | | Windows (Toolbox) | %LOCALAPPDATA%\JetBrains\Toolbox\apps\datagrip\ |
The 6 database tools will appear automatically once the plugin is installed and the MCP server is enabled.
Tools
| Tool | Parameters | Returns | |------|-----------|---------| | list_datasources | (none) | Name, driver, connection status, and write-enabled flag for each datasource | | get_schema | datasource, schema? | Tables with columns, types, primary keys, and nullability | | run_query | datasource, sql, rowLimit?, timeout? | Columns, rows, row count, and whether results were truncated | | explain_query | datasource, sql | Execution plan (auto-prefixes the correct EXPLAIN syntax for the DB driver) | | get_table_info | datasource, table, schema? | Columns, types, primary keys, foreign keys, and indexes for a single table | | search_schema | datasource, pattern | Tables and columns matching the pattern, with their types |
Datasources auto-connect on demand. Passwords must be saved in DataGrip for unattended connections.
Safety
Read-only by default. Only SELECT, EXPLAIN, SHOW, DESCRIBE, and WITH statements are allowed unless writes are explicitly enabled.
- Per-datasource write toggle: Settings > Tools > Database MCP Tools
- Query timeout: 30 seconds (default), configurable per datasource
- Row limit: 1000 rows (default), configurable per datasource and overridable per query via the
rowLimitparameter - Response size cap: ~100KB -- results are truncated if they exceed this
Development
Launch a DataGrip sandbox with the plugin loaded:
./gradlew runIde
Run tests:
./gradlew test
Source & license
This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: siddiqui-zeeshan
- Source: siddiqui-zeeshan/datagrip-database-mcp
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.