================== /// MCP /// /// ANY /// ================== [server:online] [protocol:ready]
anyquery
by julien040
Query anything (files, databases, +40 app integrations) with SQL, expose results to LLMs over the Model-Context-Protocol (MCP) and act as a MySQL-compatible server.
1.1k
63
Open Source01
anyquery
Open Anyquery’s interactive shell to run SQL queries against files, databases, or app data
02
anyquery mcp
Start the Model Context Protocol (MCP) server so LLMs can connect to Anyquery (supports --stdio or network options)
03
anyquery gpt
Generate a connection ID and facilitate linking Anyquery to LLM clients that support function calling (e.g., ChatGPT, TypingMind)
04
anyquery server
Launch Anyquery’s MySQL-compatible server so external MySQL clients can execute SQL queries
Installation
1. Prerequisites
- Go ≥1.22 installed ‑or- Docker installed.
- A PostgreSQL / MySQL / SQLite DB or SaaS account you want to query.
- API keys for cloud data sources you intend to use (GitHub PAT, Notion Secret, Salesforce OAuth, …).
2. Binaries (recommended)
# Install the CLI/server
go install github.com/julien040/anyquery/cmd/anyquery@latest
# Verify
anyquery --version
3. Docker (zero-install)
docker pull ghcr.io/julien040/anyquery:latest
docker run -d --name anyquery -p 4000:4000 \
-v $PWD/anyquery.yaml:/etc/anyquery.yaml ghcr.io/julien040/anyquery:latest
4. Configuration (anyquery.yaml)
server:
port: 4000
datasources:
github:
token: "${GITHUB_TOKEN}"
notion:
secret: "${NOTION_SECRET}"
mysql_sales:
driver: mysql
dsn: "user:pass@tcp(mysql:3306)/sales"
Export the required secrets before starting:
export GITHUB_TOKEN=ghp_***
export NOTION_SECRET=secret_***
5. Run locally
anyquery serve --config ./anyquery.yaml
6. Test
curl -X POST http://localhost:4000/v1/execute \
-H 'Content-Type: application/json' \
-d '{"query":"SELECT * FROM github.repos LIMIT 10;"}'
Documentation
License: Other
Updated 7/30/2025