DB Connect

One database client for your entire backend stack.

MySQL + PostgreSQL + Amazon Redshift + DynamoDB + Redis + MongoDB + Elasticsearch/OpenSearch — one fast, free, native desktop IDE instead of seven separate tools and seven separate sets of saved credentials.

Guides: MySQL · PostgreSQL · Redshift · DynamoDB · Redis · MongoDB · Elasticsearch / OpenSearch · Case study: MySQL CPU 90% → 30% Compared with: DataGrip · DBeaver · TablePlus · MySQL Workbench · Sequel Ace · pgAdmin · RedisInsight · NoSQL Workbench · Dynobase · MongoDB Compass · Studio 3T

Download

| | | |---|---| | macOS (recommended: Homebrew) | brew install --cask shubhesh07/db-connect/db-connect | | macOS (no Homebrew) | curl -fsSL https://github.com/shubhesh07/db-connect/releases/latest/download/install.sh \| bash | | Windows | Download DBConnect-Windows-Setup.exe |

Full manual-download links (DMG, portable ZIPs) are in Manual downloads below.

Why DB Connect?

• One tool, not three. If your stack is MySQL or PostgreSQL for transactional data, Redshift for analytics, DynamoDB for high-throughput key-value access, Redis for caching, MongoDB for documents and Elasticsearch or OpenSearch for search, you know the pain of a GUI client, credential store and query history per engine. DB Connect treats all seven engines as equal, first-class citizens.
• Free forever. No subscription, no trial limits, no feature paywall.
• Fast. Opens in under 2 seconds. ~35 MB installed. Built with Go + Wails — a native app, not Electron, so there's no bundled Chromium sitting in your dock for a database browser.
• DynamoDB done properly. A visual Scan/Query/GetItem builder and a PartiQL editor for writing real SELECT * FROM "table" WHERE ... statements — most GUI database tools ignore DynamoDB entirely or bolt on a bare-bones table browser.
• Built to work with your AI coding tools. A local MCP server lets Claude Code, Cursor, or any MCP-aware client query your already-open connections directly — schema, read-only queries, gated behind a per-install auth token, an explicit connection allow-list, and time-bounded write grants.
• No cloud sync, no telemetry. Credentials are AES-256-GCM encrypted with the key in your OS keychain. Nothing you connect to or query ever leaves your machine.

MySQL · Redshift · DynamoDB · Elasticsearch

| | MySQL | Redshift | DynamoDB | Elasticsearch / OpenSearch | |---|---|---|---|---| | SQL query editor | ✅ | ✅ | — | ✅ (_sql / _plugins/_sql) + Kibana-style console | | PartiQL query editor | — | — | ✅ | — | | Visual Scan/Query/GetItem builder | — | — | ✅ | — | | Schema/table browser | ✅ | ✅ | ✅ | ✅ (indices, aliases, mapping fields) | | Context-aware autocomplete | ✅ | ✅ | ✅ (keywords, tables, per-table attributes) | ✅ (endpoints, query DSL, mapped fields) | | EXPLAIN visualization | ✅ | ✅ | — | ✅ (Profile and Explain trees) | | Inline row editing | ✅ | ✅ | ✅ (item editor) | ✅ (grid + document editor) | | Table designer (CREATE/ALTER) | ✅ | — | — | — | | Index manager | ✅ | — | — | ✅ (new/delete index, aliases, reindex) | | Transactions | ✅ | — | — | — | | SSH tunnel | ✅ | — | — | ✅ | | GSI-aware querying | — | — | ✅ | — | | Multiple AWS auth modes (access key, SSO profile, IAM role) | — | — | ✅ | — | | Real backup/restore (mysqldump) | ✅ | ✅ | — | — | | Native performance monitoring | ✅ | ✅ | ✅ (capacity/throttle, no CloudWatch) | ✅ (health, rates, tasks, shards, hot threads) | | Export (CSV, JSON, Excel) | ✅ | ✅ | ✅ | ✅ (streams through cursor or scroll) |

Elasticsearch / OpenSearch

New in v3.7.0, the seventh engine. Full guide.

• Connect to Elasticsearch or OpenSearch: self-managed, Elastic Cloud, AWS OpenSearch Service and AWS Elasticsearch domains. Username and password or an API key, TLS with a custom CA, SSH tunnel. The distribution and version are detected on connect; AWS domains get a port and TLS hint if misconfigured.
• Explorer: indices with aliases, doc counts and health, expanding to mapping fields (nested objects, multi-fields). Right-click for Query (SQL), Search (console), Mapping, Settings, Stats, Count, Refresh, Reindex, Add alias, New index, Delete index, New document and Bulk insert; cluster-level Health, Nodes, Shards and Aliases.
• SQL and console in one editor: SELECT … FROM index (Elastic _sql, OpenSearch _plugins/_sql, or the Open Distro endpoint on AWS Elasticsearch domains, chosen automatically) or Kibana-style requests such as GET /index/_search with a JSON body, several per run, with endpoint and query-DSL autocomplete that knows your mapped field names.
• Results and document editing: hits become rows with _index, _id, _score and the source fields; aggregations arrive as a second result set; _cat output becomes a table. Edit cells, delete rows and insert documents from the grid (one _update per document) or in the full document editor. Load more through the SQL cursor or from/size; export streams through the cursor or scroll API.
• Profile and Explain: per-shard query, collector and aggregation timings as a tree, or why each hit scored.
• Performance pane: cluster health, search and indexing rates and latencies per node and index, heap, CPU and rejections, plus Tasks (with cancel), Shards (with allocation explain), Hot threads, Pending tasks and circuit breakers.
• Safety: read-only profiles and production confirmation apply to every write, including DELETE, PUT, _update_by_query, _delete_by_query, task cancel and document edits.
• MCP: list_tables, describe_table, run_query (SQL or console), explain_query, active_sessions and cluster_health work on Elasticsearch connections; writes still need a grant.

How it compares

| Feature | DB Connect | DBeaver (Community) | DataGrip | TablePlus | Beekeeper Studio (Community) | |---|:---:|:---:|:---:|:---:|:---:| | Price | Free | Free | Free (non-commercial) / ~$99 yr¹ (commercial) | $99+ one-time² | Free | | MySQL | ✅ | ✅ | ✅ | ✅ | ✅ | | Redshift | ✅ | ✅ | ✅ | ✅ | ✅ | | DynamoDB | ✅ (PartiQL + visual builder) | — (paid tiers only) | — | — | — (paid tiers only)³ | | Native app (not JVM/Electron) | ✅ | — | — | ✅ | — | | Install size | ~18MB | ~400MB | ~800MB | ~80MB | ~500MB | | Startup time | <2s | 5–15s | 10–30s | <3s | — (Electron) | | SSH tunnel | ✅ | ✅ | ✅ | ✅ | ✅ | | Query history (persists across restarts) | ✅ | — (session-only in Community) | ✅ | ✅ | ✅ | | Real backup/restore | ✅ | — | — | — | ✅ | | Native performance monitoring | ✅ | — | — | — | — | | Local MCP server (AI tool integration) | ✅ | — | — | — | — | | macOS | ✅ | ✅ | ✅ | ✅ | ✅ | | Windows | ✅ | ✅ | ✅ | ✅ | ✅ |

¹ DataGrip has been free for non-commercial use (learning, hobby, open-source) since October 2025; commercial use still requires a paid license. ² TablePlus's free tier caps open tabs/windows/filters; the one-time license runs $99–$129 depending on device count. ³ Beekeeper Studio's Community edition includes Redshift but gates DynamoDB to its paid Indie/Professional/Business tiers.

Comparison reflects each product's free/community tier where one exists, verified against each vendor's own docs as of this writing. Corrections welcome — open an issue if something's out of date.

vs. popular MySQL-only tools

DB Connect, DBeaver, DataGrip, TablePlus, and Beekeeper Studio above all support multiple database engines. MySQL Workbench and Sequel Ace are two of the most popular MySQL-only clients, so they get their own table instead of a row full of dashes above:

| Feature | DB Connect | MySQL Workbench | Sequel Ace | |---|:---:|:---:|:---:| | Price | Free | Free (Community) | Free (open source, MIT) | | MySQL | ✅ | ✅ | ✅ | | Redshift | ✅ | — | — | | DynamoDB | ✅ | — | — | | Native app | ✅ | ✅ | ✅ | | Platforms | macOS, Windows | macOS, Windows, Linux | macOS only | | SSH tunnel | ✅ | ✅ | ✅ | | Query history | ✅ | ✅ (persists across restarts) | ✅ | | EXPLAIN | ✅ Visual | ✅ Visual | ✅ (text output) | | Backup/restore | ✅ | ✅ | ✅ (SQL dump-based) |

MySQL Workbench is Oracle's official free tool and a genuinely solid choice if MySQL is the only engine you touch — its Visual Explain is long-standing and well-regarded. Sequel Ace (the free, open-source successor to Sequel Pro) is excellent if you're macOS-only and MySQL-only. Neither reaches beyond MySQL/MariaDB, which is the gap DB Connect fills if Redshift or DynamoDB are also in your stack.

30-second quick start

brew install --cask shubhesh07/db-connect/db-connect

1. Launch DB Connect
2. Add Connection → choose MySQL, PostgreSQL, Redshift, DynamoDB, Redis, MongoDB, or Elasticsearch/OpenSearch
3. Enter credentials → Test Connection → Save
4. Write a query (or, for DynamoDB, use the visual builder or the PartiQL editor) → Cmd+Enter to run

Connect Claude (MCP) in three steps

1. In DB Connect: sidebar → MCP Server → enable the server and tick the connections Claude may see (the allow-list).
2. In the same dialog, press copy next to Set up in Claude — it copies a ready-made command with your URL and auth token filled in:

   claude mcp add --transport http db-connect http://127.0.0.1:47823/mcp --header "Authorization: Bearer <your-token>"
   

3. Paste it in your terminal. Done — Claude Code can now list databases/tables/collections/indices, describe schemas, and run read-only queries (SQL, or console requests on Elasticsearch/OpenSearch) against the allow-listed connections. (/mcp inside Claude Code shows the connection.)

For Claude Desktop (or any JSON-configured MCP client), the dialog also has a copy button for the matching mcpServers JSON block — merge it into claude_desktop_config.json (Settings → Developer → Edit Config) and restart Claude Desktop.

Writes stay blocked until you grant them per connection, time-bounded, from the desktop app — an agent can never grant itself write access.

Keyboard shortcuts

| Shortcut | Action | |---|---| | Cmd+Enter | Execute query | | Cmd+D | Select current statement | | Cmd+K | Command palette | | Cmd+T / Cmd+W | New tab / close tab | | Cmd+S | Save query | | Cmd+Shift+F | Format SQL | | `Cmd+E