Your databases.
Inside your
editor.

DBDeck is a free, open source database client for VS Code and Cursor. Browse tables, run queries and edit rows next to your code. No account, no paywall, no telemetry.

Free foreverMIT licensedNothing leaves your machine
orders · shop · postgres
wherestatus in ('paid', 'shipped')
idcustomerccstatustotal
1187Mia NovakCZpaid182.90
1186Oliver BergSEshipped74.90
1185Ava JensenDKdelivered311.90
1184Emma DuboisFRrefunded48.90
1183Lucas SilvaPTpaid126.90
1182Anna KowalskaPLpending205.90
1181Zoe KimKRdelivered89.90
1–7 of 1,200 · 4 ms connected
Fig. 01 — the data grid, mid-shift
vs code · dbdeck
DBDeck in VS Code: connection tree with PostgreSQL, Redis and MongoDB, and the customers table in the data grid
Fig. 02 — connections on the left, a 240-row table on the right

Eight services. One sidebar.

Keep every connection in one tree, grouped the way you work. SSH tunnels, SSL and read-only mode come with each of them.

  • SSH tunnels with password or private key
  • Read-only connections block every write
  • Docker containers become connections in one click
add connection
  • PostgreSQLSQL, grid editing, DDL, diagrams
  • MySQL / MariaDBSQL, grid editing, DDL, diagrams
  • ClickHouseSQL editor, data browsing, DDL
  • MongoDBDocuments, filters, aggregation
  • RedisKey tree, value editors, TTL, CLI
  • ElasticsearchIndices, documents, console
  • S3 / MinIO / R2Buckets, uploads, downloads
  • DockerContainers, logs, shell, discovery
Fig. 03 — every type DBDeck connects to

Write SQL where you write code.

Put the cursor in a statement and press ⌘ Enter. Each statement gets its own result tab, with a row filter, JSON view and CSV or JSON export.

  • Table and column completion with aliases
  • Run one statement or the whole file
  • Inline edits saved in one transaction
query · shop · postgres
SQL editor with a join query and the Query Results panel listing revenue per customer
Fig. 04 — 12 rows in 59 ms
public · diagram
Schema diagram of seven tables with primary keys, foreign keys and relationship lines
Fig. 05 — 7 tables, 7 foreign key links

The schema, drawn for you.

Open any PostgreSQL or MySQL schema as a diagram. Foreign keys link column to column. Drag tables, pan, zoom and search; the layout is remembered.

  • Primary and foreign keys marked
  • Arrange and Fit in one click
  • ClickHouse tables too

Not only SQL. Real editors for keys and documents.

Redis keys grouped by separator with editors for every type. MongoDB documents with filters and shell-style scripts. Elasticsearch with a Kibana-style console.

  • Hash, list, set, zset, stream, RedisJSON
  • db.users.find({...}).sort(...) just works
  • S3 buckets: open, upload, copy s3:// URI
user:42 · cache · redis
Redis key tree grouped by prefix and the hash editor for user:42
Fig. 06 — a hash, four fields, no expiry

Your credentials never leave your machine.

DBDeck talks to the databases you add and to nothing else. There is no backend to sign in to, so there is no backend to leak.

  • AI queries share table names, never rows
  • Bring ChatGPT, your own API key or Ollama
  • Destructive actions always ask first
what dbdeck keeps
datalives innote
Connection settingsEditor global statethis machine
Passwords and keysOS keychain (SecretStorage)encrypted
Remember password offSession memorygone on reload
Query resultsNot stored—
TelemetryNone—
No telemetry, analytics or cloud sync. Read the code to check.
Fig. 07 — the full list

Questions, answered.

Free. No account. No paywall, ever.

Install it, add a connection, open a table. You are querying in under a minute.