WhoDB
A lightweight, self-hosted workspace for your databases, with browsing, editing, querying, and AI-powered natural language interactions.
Evidence: README mentions encrypted sessions and WHODB_ENCRYPTION_KEY, indicating some sensitive data handling; CI workflows use step-security/harden-runner to restrict network egress, showing least-privilege awareness; but no user confirmation mechanism, data flow transparency, rollback mechanism, or clear source attribution found. Deductions: missing user confirmation, data flow transparency, rollback, and source attribution evidence.
Evidence: README and CI workflows are consistent, described features align with code structure; dependencies managed via go.mod and package.json, but no dependency availability guarantee; failure messages not clearly documented. Deductions: unclear failure messages, dependency availability unverified.
Evidence: README clearly identifies target users (database developers, team collaboration) and multiple usage scenarios (local, desktop, CLI); supports many databases with capability boundaries explained; but trigger precision (e.g., AI feature triggers) not detailed; environment fit (Docker, desktop, CLI) documented. Deductions: insufficient trigger precision.
Evidence: README well-structured with quick start, installation options, AI provider configuration; install notes detailed; naming stable (WhoDB); examples and FAQ links present; known limitations not explicitly listed; license is Apache-2.0; versioning and changelog not reflected in files; maintenance responsibility has support email and contribution guide. Deductions: missing known limitations and version/changelog.
Evidence: High output usability (multiple interfaces and CLI); clear marginal value (one-stop database management); reasonable cost-benefit (open-source free). Deductions: no specific performance or cost data to support.
Evidence: README claims align with code structure, but no traceable verification methods; cross-source corroboration insufficient (only README and CI); facts and inferences not clearly separated. Deductions: lack of traceability and cross-source verification.
- No user confirmation mechanism found; confirm before sensitive operations.
- Data flow transparency insufficient; review network requests and data handling.
- No rollback mechanism; backup data before operations.
- Dependency security not fully verified; check for vulnerabilities.
What does this agent do, and when should you use it?
WhoDB is a self-hosted database workspace available as a web app, desktop application, and CLI. It supports a wide range of databases including PostgreSQL, MySQL, SQLite, MongoDB, Redis, and more. Key features include a spreadsheet-style data grid for browsing and editing rows, an interactive schema graph, a multi-cell query scratchpad, import/export and mock data generation. Optional AI integration via providers like Ollama, OpenAI, and Anthropic allows users to ask questions in plain English. WhoDB also offers a CLI and MCP server for terminal-based access and AI tool integration. The community edition is Apache-2.0 licensed, with a commercial WhoDB Platform for organizations.
WhoDB provides an end-to-end workflow for database interaction. Users connect via browser, desktop, or CLI, then browse, filter, sort, and edit rows in a data grid. The schema graph visualizes table relationships. The scratchpad allows writing and executing SQL queries with autocomplete and history. Users can import/export data and generate mock data for development. The chat interface allows connecting to AI providers (OpenAI, Anthropic, Ollama, LM Studio) to ask natural language questions about the data. The backend is Go, frontend is React/TypeScript, and deployment via Docker is supported with environment variables for encryption keys and AI provider configuration.
- A database administrator quickly inspects a local or remote database's schema and rows without installing a heavy client.
- A developer generates mock data or imports/exports data to assist with development and testing.
- A data analyst asks natural language questions and gets query results without writing SQL.
- A new team member explores an unfamiliar database through the schema graph to understand relationships.
- A sysadmin runs queries and manages databases on a headless server using the CLI or MCP server.
- An organization evaluates WhoDB Platform for shared projects, SSO, audit logs, and other enterprise features.
What are this agent's strengths and limitations?
- Supports a broad range of databases including PostgreSQL, MySQL, SQLite, MongoDB, and Redis, covering common use cases.
- Visual schema graph and data grid editing lower the barrier to database interaction.
- Flexible AI integration with multiple providers and no backend configuration required for hosted providers.
- Multiple deployment options: self-hosted Docker, desktop apps, and CLI, suitable for various environments.
- MCP server enables integration with AI tools for automation.
- Community edition features may vary per database; advanced features like SSO and audit logs are only in the paid Platform.
- AI features depend on external providers (Ollama, OpenAI, etc.), requiring network access or local services.
- The browser-based data grid may struggle with very large datasets.
- Configuring custom AI providers via environment variables may be intimidating for non-technical users.
How do you install or deploy this agent?
Run with Docker:
docker run --rm -it -p 8080:8080 clidey/whodbTo persist sessions, first generate an encryption key:
openssl rand -hex 32Then mount /data and set the environment variable:
docker run -it -p 8080:8080 \
-v whodb-data:/data \
-e WHODB_ENCRYPTION_KEY=your_saved_64_character_hex_key \
clidey/whodbAlternatively, install the desktop app (macOS/Windows/Snap) or the CLI.
How do you use this agent?
After starting, open http://localhost:8080 in your browser and enter your database connection details. You can browse and edit data in the grid, explore the schema graph, and write queries in the scratchpad. To enable AI, open Chat, click 'Add Provider', and enter an API key for OpenAI/Anthropic, or run local Ollama/LM Studio. For CLI, run whodb-cli to open the terminal UI, or whodb-cli mcp serve to start the MCP server.
How does this agent compare with similar options?
WhoDB can be seen as an alternative to traditional database clients like TablePlus or DBeaver, or internal tool builders like Retool, but with a lighter footprint and a focus on self-hosting and AI integration.