Quarry
Allows querying and managing MariaDB databases with support for read-only mode, safety rails, and structured results.
Allows querying and managing MySQL databases with support for read-only mode, safety rails, and structured results.
Allows querying and managing AWS Neptune graph databases with support for read-only mode, safety rails, and structured results.
Allows querying and managing PostgreSQL databases with support for read-only mode, safety rails, and structured results.
Allows querying and managing Redis databases with support for read-only mode, safety rails, and structured results.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
Quarry 2
A native database workbench: one reusable Rust kernel, CLI, browser GUI and MCP. PostgreSQL, MySQL, SQLite, Redis and experimental Neptune; explicit query/write policies, SSH tunnels, saved queries, local Docker services and a SQL assistant.
Install
Quarry 2.x ships as a native Rust executable.
curl -fsSL https://yiminspace.github.io/quarry/install.sh | sh
qy --version
qy guiOr download a native archive and SHA256SUMS from GitHub Releases. The native executable needs neither Python nor a Rust toolchain. Supported: macOS/Linux arm64/x86_64; Linux glibc 2.28+. See install and upgrade.
Installer-managed copies can download new stable versions in the background.
The next launch activates the verified update. Run qy update to update now.
Related MCP server: sqlite-mcp
Use
Run qy gui stop from another terminal to shut down the local standalone GUI.
Use the same --config, --workspace, --host, and --port overrides as at
launch. If you started with --port 0, use the port in the printed GUI URL.
qy --workspace /path/to/workspace connections
qy --workspace /path/to/workspace exec shop --env dev --sql 'SELECT 1'
qy --workspace /path/to/workspace mcpWorkspaces contain connections.toml and reusable SQL query files. Actual
credentials remain local. GUI queries stay read-only; CLI writes require
--write and production confirmation. MCP writes require both server and
request authorization. SQLite opens existing files; it never invents a database
when a connection path is wrong. Assistant proposals never execute themselves.
Ordinary queries use native database drivers. SSH needs system ssh; Local services need Docker, and PostgreSQL schema sync needs pg_dump/psql.
Rust and Python
Rust owns main and all 2.x distribution. quarry-core, quarry-runtime,
quarry-local, quarry-assistant and quarry-service can be embedded directly
in host applications. Architecture.
For a hosted workbench, the official Docker image serves the GUI, authentication,
read-only API and execution from one configuration file. Use a built-in
administrator or configure an existing platform's IAM. Standalone qy runs
locally with its own workspaces.
See the hosting guide for deployment, access grants and offline
behavior. The Rust SDK remains available for advanced embedding.
Python 1.11.0 is preserved at its tag and the protected archive/python-1.x
branch. Python import APIs do not ship in 2.x. The Python source retained in this
checkout is a development/test baseline, not a runtime dependency or published
Python implementation. Existing configuration, saved queries and histories are
read by Rust. See upgrade instructions before changing installer ownership.
Build and verify
For frontend development, start the Rust workbench and Vite in separate terminals.
qy gui --no-browsercd web
npm ci
npm run devOpen the Vite URL ending in /quarry/. Vite reloads frontend changes and forwards
/api to the local Rust workbench on port 8765. If qy gui uses a different
port, start Vite with QUARRY_DEV_API_PORT=<port> npm run dev.
For a native build and the existing test workflow:
cd web && npm ci && npm run build && cd ..
cargo build --release --locked -p quarry-cli
cargo test --workspace --all-targets --lockedGUI feature matrix · Compatibility · Verification boundaries · 中文
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Draxlr's remote MCP server connects AI assistants to your SQL databases and dashboards. Explore schemas, run read-only queries, manage saved queries and dashboards, and export results, all with row-level security so each user sees only their own data.
The BigQuery remote MCP server is a fully managed service that uses the Model Context Protocol to connect AI applications and LLMs to BigQuery data sources. It provides secure, standardized tools for AI agents to list datasets and tables, retrieve schemas, generate and execute SQL queries through natural language, and analyze data—enabling direct access to enterprise analytics data without requiring manual SQL coding.
Hosted SQL Server MCP server for Claude, ChatGPT and Cursor, with each action switched per account.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceAn MCP server that enables AI assistants to query and interact with SQLite databases through natural language. It includes built-in security guardrails such as PII redaction, SQL injection blocking, and query rate limiting.-
- AlicenseNot gradedqualityDmaintenanceAn MCP server that enables AI agents to interact with SQLite databases by querying schemas, executing SQL, and inspecting table metadata. It supports safe database access through configurable read-only modes, query timeouts, and dry-run execution plans.MIT
- AlicenseAqualityCmaintenanceAn MCP server that gives AI agents access to configured databases (PostgreSQL, MySQL, Redshift, SQL Server) with SSH/AWS SSM tunnels, pluggable secret providers, and strict per-instance isolation.1080 npmMIT
- AlicenseNot gradedqualityDmaintenanceOctoQuery is an MCP server that converts databases into AI-accessible tools by exposing each configured database as a tool that accepts SQL queries and returns JSON results, enabling AI agents to query your data directly.3MIT