Connecting to a database

RED connects to six database families:

Each connection opens a session owned by RED's backend service, which keeps it alive between queries so the schema tree and result grid stay responsive.

Opening a connection

Add one with ⌘NCtrl+N from the welcome screen. The form is driven by the engine you pick, so a Redis connection asks for what Redis needs and a SQLite one asks for a file. Leave the host blank and it falls back to localhost, the way psql and redis-cli do.

Passwords are stored in your operating system's keychain (Keychain / Credential Manager / Secret Service), never in plaintext.

Once connected, the schema explorer in the sidebar lists namespaces, tables, views and other objects - select one to browse it, or start querying from a SQL editor tab.

Connection options

• Read-only - the driver opens read-only and every write path is refused up front, in the UI and for the AI assistant alike.
• Environment - mark a connection Local, Dev, Staging or Prod. The write guards follow the marker, and the connection card carries a coloured badge. See Safety and guards.
• TLS - a checkbox that dials rediss:// for Redis, requires SSL on MySQL, and uses HTTPS for ClickHouse. Pasting a DSN that already says so (an sslmode=require, a rediss:// URL) sets it for you.
• AI access - override the global assistant switch and tier per connection, so a production database can be pinned to off or schema.

SSH tunneling

A connection can reach a database through an SSH jump host - point RED at the bastion and it opens the tunnel for you, the same way a DataGrip jump host works. Key and passphrase secrets go to the keychain with the rest.

Proxies

"Connect via proxy" reaches a database through a SOCKS5 or HTTP CONNECT proxy, with optional auth whose password is stored in the keychain. A connection uses either a proxy or an SSH tunnel, not both.

Redis Sentinel and cluster

A Redis connection can name a Sentinel master group, in which case the host and port point at a Sentinel and RED resolves the current master itself. Cluster deployments are scanned across slots, so a key search fans out over the nodes.

Import existing connections

Already using another tool? The import wizard reads saved connections from:

• DBeaver
• DBGate
• JetBrains DataGrip / IntelliJ (dataSources.xml)
• RedisInsight
• Plain credential files: ~/.pgpass, ~/.my.cnf, ~/.pg_service.conf

It decrypts what it can and moves the credentials into your OS keychain; anything unrecoverable is flagged for re-entry rather than silently dropped. Redis and MongoDB entries are imported alongside the SQL ones. Look for the import link on the welcome screen or run it from the command palette (⌘KCtrl+K).

Multiple connections

You can hold multiple connections at once and switch between them without losing query state. Switching is keyboard-first:

The welcome screen sorts and filters the roster by engine and environment (several values at once), remembers those filters between launches, and pages through large rosters with ← / →. Press / to jump into the search box, which also matches engine and environment.

Restore last session reopens the tabs you left a connection with: their SQL, focus, pins, per-tab database, browsed table and pane layout.

Database context

MySQL and ClickHouse connections browse a whole server rather than one database. The editor breadcrumb reads connection / database / table, where the database is a picker showing what an unqualified FROM users resolves against. The target is per tab, so a split view can compare two databases over one connection.

Production-safe by design

RED is built for inspection: it shows roughly enough data to make decisions quickly without loading entire tables into memory. The result grid is windowed over the full result set using keyset pagination, so browsing a multi-million row table stays fast and bounded.