SuperAPI AI Directory
Partner with usδΈ­ζ–‡AI newsFavorites
All toolsAI codingdbconvert-streams-public

dbconvert-streams-public

πŸŽ–οΈ 🏎️ 🏠 🍎 πŸͺŸ 🐧 - Read-only SQL across PostgreSQL, MySQL, S3-compatible storage and Parquet/CSV/JSON files, with cross-source joins federated by DuckDB.

Open website

About this tool

DBConvert Streams is a database IDE that integrates data querying, migration, and real-time CDC (change data capture) into the same workflow. Its goal is to solve the problem in traditional solutions where database clients, migration scripts, and CDC pipelines are fragmented from one another.

Main uses

  • Explore data structures in databases, local files, and S3 storage
  • Run read-only SQL queries across multiple data sources, with federated query support
  • Quickly migrate large amounts of data between databases
  • Synchronize INSERT, UPDATE, and DELETE changes in real time
  • Let AI assistants read real-time schemas, data, and stream status through MCP, providing query suggestions and troubleshooting

Core capabilities

Database IDE and workspace (free)

  • Data explorer: browse databases, files, and S3 in one interface
  • ER diagram: visualize database relationships
  • Schema comparison: compare schemas and sample data across databases
  • Schema navigation: preserve state across connections and support search
  • Federated SQL: query multiple databases and file sources at the same time, for example joining MySQL, PostgreSQL tables and CSV/Parquet files, without intermediate exports

Built-in AI chat (new in 2.5.0)

Ask questions about database workspaces already open in the desktop app (tables, views, files, SQL console, connections, or migration/CDC streams). AI Chat starts with live workspace context and can inspect schemas and diagnose stream status. It supports using installed agent CLIs (Claude Code, Codex, GitHub Copilot CLI, or OpenCode), automatically provides a read-only subset of tools, and cannot change connections, configuration, streams, or data.

AI assistants via MCP (new in 2.4)

A built-in MCP server exposes 27 read-only tools to Claude, Cursor, VS Code Copilot, Windsurf, Gemini CLI, and Codex, including:

  • Connections and schemas: list connections, databases, schemas, tables, views
  • Table and view inspection: describe tables/views, preview data, run and explain SELECT
  • Schema and data comparison: compare schemas, compare data samples
  • Stream diagnostics: list streams, inspect streams, get stream status, throughput statistics, recent errors, and logs
  • Files and S3: list workspace files, get file schemas, preview files, list S3 buckets andObjects
  • Federated queries: Run and interpret cross-source read-only queries

Server-side filters only allow SELECT; AI clients cannot modify connections, configurations, streams, or data.

Data Migration (Load Mode)

Quickly move large datasets between databases, with automatic schema conversion and validation. Performance example: migrating 23 million rows (4.38 GB) from MySQL to Parquet takes only 35.7 seconds, at a rate of 136 MB/s.

Real-Time CDC

Stream INSERT, UPDATE, and DELETE operations from the source to the target in real time with minimal latency, supporting targets such as databases, files, and S3 storage.

How to Use

  • Desktop App: Download the Windows, macOS, or Linux version from the official website, no account required.
  • Docker Self-Hosting: Can be deployed on a local server, VPS, or your own infrastructure.
  • AI Client Extension: Download the .mcpb file and drag it into Claude's extension settings, then enter the connection string and folder. The same server can also run as a container for any MCP client that can start containers; Cursor and VS Code support one-click addition.
  • API: The UI covers connections, table selection, federated SQL, stream configuration, and monitoring; the API is for users who need scripted deployment or CI/CD integration.

Use Cases

  • Need to query and explore across multiple databases and file formats
  • Want to complete migration and continuous synchronization within one tool
  • Want to let AI assistants safely read real-time data context to assist diagnosis
  • Environments that do not need cloud dependencies and want to avoid vendor lock-in

Limitations

  • Not suitable for scenarios requiring 100+ connectors (SaaS, APIs, etc.)
  • Not suitable for scenarios already running Kafka pipelines at scale
  • Not suitable for scenarios requiring complex ETL / transformation
  • AI tools are read-only and cannot perform write operations
  • The product itself is proprietary software; the MIT license in the repository only covers examples, documentation, and assets
  • The macOS version of the MCP extension binary is planned for a future release

Supported Data Sources and Targets

Sources: MySQL / MariaDB / Percona, PostgreSQL / CockroachDB, Amazon RDS, Aurora, Google Cloud SQL, Azure Database, local files (CSV, JSONL, Parquet), S3-compatible storage (AWS S3, MinIO, DigitalOcean Spaces, Wasabi) Targets: MySQL / PostgreSQL, CSV / JSONL / Parquet (local files), Amazon S3 / MinIO / S3-compatible storage, Google Cloud Storage, Azure Blob Storage

The database IDE is permanently free. For pricing details on data migration and CDC streaming, refer to the official pricing page.

Related tools

Browse category