FAQ
Answers to the most common questions about Datavor — setup, compatibility, security, and how it works.
General
What is MCP?
MCP (Model Context Protocol) is an open standard created by Anthropic that lets AI agents like Claude connect to external tools and services. Datavor is an MCP server — meaning Claude can call it directly, like a skilled database assistant built into your conversation.
Do I need to know SQL?
No. Just describe what you want in plain English to Claude. Datavor translates your request into precise SQL operations automatically — no database expertise required. For advanced users who want to run custom queries, that option is also available via the execute_query tool.
What databases are supported?
Phase 1 supports MySQL and PostgreSQL. Phase 2 will add MongoDB, SQLite, Supabase, AWS RDS, Google Cloud SQL, and more. Both directions are supported — MySQL → PostgreSQL and PostgreSQL → MySQL.
Can it sync bidirectionally?
Yes. You specify the source and target in each request. MySQL → PostgreSQL and PostgreSQL → MySQL both work. Datavor handles all the data type conversions between the two databases automatically.
Installation
What operating systems are supported?
Phase 1 supports macOS (10.15 Catalina or later). Windows and Linux support is coming in Phase 2. Node.js v18 or higher is required on all platforms.
Does it work with remote databases?
Yes. Datavor connects to any accessible MySQL or PostgreSQL server — localhost, your office network, or cloud-hosted instances (AWS RDS, DigitalOcean, Supabase, etc.). Just provide the host, port, username, and password when connecting.
Claude doesn't see Datavor after install. Why?
Two things to check:
claude_desktop_config.json, you must fully quit Claude Desktop (Cmd+Q) and reopen it. Just closing the window is not enough.Security & Privacy
Is my data sent to any third party?
No. Datavor runs locally on your machine and connects directly to your databases. Your data never passes through Datavor's servers — all operations happen between Claude (running on your device via Claude Desktop) and your databases directly.
How are my database credentials handled?
Credentials are passed directly to your database connection within the session. They are not stored on disk by Datavor, not logged, and not transmitted anywhere outside your local machine. Each session starts fresh.
Usage
Do I need to reconnect every time?
Yes — connections are session-scoped, meaning they exist for the duration of your Claude Desktop conversation. When you start a new conversation, just ask Claude to reconnect to your databases. This is a deliberate security design so credentials are never stored persistently.
What happens if a sync fails mid-way?
Datavor syncs in batches and logs exactly which rows succeeded or failed. A partial failure will not corrupt your target database — each batch is atomic. You'll receive a full report showing how many rows succeeded and which failed with the reason why. You can safely retry.
Can it handle large tables?
Yes. Datavor uses batch processing (default 1,000 rows per batch) to handle tables of any size without memory issues. For very large tables, incremental sync is strongly recommended after the first full sync — it can be up to 98% faster by only transferring changed rows.
Pricing
Is Datavor free?
Phase 1 is completely free — no credit card, no time limit, no usage cap. Phase 2 will introduce paid tiers to support ongoing development. Users who join the waitlist will receive early access and preferential pricing.
What will Phase 2 cost?
Pricing for Phase 2 has not been finalised yet. The goal is to keep Datavor accessible — a free tier will remain available. Join the waitlist on the homepage to be notified when pricing is announced.