¶Database

e includes a Database panel for browsing and querying your project's databases without leaving the editor. It supports MySQL/MariaDB, PostgreSQL, SQLite and ClickHouse (over its HTTP interface), with an optional SSH tunnel for remote databases.

Toggle it with ⌘3 (or the command palette → Toggle Database Panel). The panel sits on the right by default; move it to the left under Settings → Database panel.

¶Adding a connection

Click ＋ in the panel header:

• From .env — reads DB_CONNECTION, DB_HOST, DB_DATABASE, DB_USERNAME, DB_PASSWORD (and SQLite's DB_DATABASE path) from the project's .env, following Laravel conventions.
• Manually — pick an engine and fill in host / port / database / user / password, or a file path for SQLite.

For remote databases, tick Use SSH tunnel and fill in the SSH host, user and either a private key or a password — e forwards a local port through the system ssh for the lifetime of the connection.

Connections are saved per project in ~/.config/e/databases.json — never written into the project folder, so nothing can be committed.

¶Where passwords are kept

Database passwords, SSH passwords and SSH key passphrases are not stored in databases.json. They go to the operating system's credential store:

databases.json holds only the non-secret fields plus a reference, and is written 0600 (readable by your user only). If no credential store is reachable — a headless Linux box with no Secret Service, say — e falls back to keeping the secrets in that file, still 0600, and says so on stderr.

Connections saved by an earlier version keep working: their secrets move to the credential store the next time the connection list is saved, and the file's permissions are tightened as soon as e reads it.

When adding manually you can press Test to verify the connection before saving.

¶Browsing & querying

• Click a connection to connect and list its tables.
• Use the filter box to narrow the table list.
• Click a table to open its rows in the results grid (200 per page).
  • Switch between Data and Structure. Structure lists columns (types, nullability, keys — primary keys are marked 🔑) and the table's indexes (name, UNIQUE/INDEX, and the columns each covers).
  • Click a column header to sort (ascending → descending → off).
  • Page through rows with ‹ Prev / Next ›, or pan with the arrow keys.
  • 💾 saves the current SQL as a named query; Saved ▾ loads it back.
  • ⬇ CSV exports the current result to a file.
• Click ⌗ on a connection to open a blank query editor. Type SQL and press ⌘↵ (or Run) to execute. SELECT/SHOW/EXPLAIN show a grid; other statements report the number of affected rows.

Results are capped at 1000 rows.

¶Editing data

In the Data view of a table that has a primary key:

• Double-click a cell to open the edit dialog. Change the value (or tick NULL) and press ⌘↵. Edits are staged, not written immediately — see Environments & safety for Submit/Revert.
• The dialog also offers:
  • Delete row — stages a deletion (requires a primary key).
  • Follow FK → — if the column is a foreign key, jumps to the referenced table filtered to the linked value.
  • Related → — shows rows in other tables that reference this row (reverse foreign keys).
  • Filter to value — restricts the current table to rows matching that cell (WHERE col = value / IS NULL), composing with sort and pagination. An active filter appears as a chip in the toolbar — click it to clear.
• + Row on the toolbar opens an insert dialog with one field per column (each with a NULL toggle). Columns left blank (and not marked NULL) are omitted, so database defaults and auto-increment apply.

A connection that looks like production defaults to read-only (a 🔒 / 🔓 badge toggles it); see below.

¶Inline SQL in PHP

SQL inside PHP strings (DB::select(...), ->whereRaw(...), and similar) is syntax-highlighted. With the cursor inside such a string and a database connected:

• ⌘↵ — Run SQL under cursor: runs the query and shows the result grid.
• ⌥⌘↵ — Explain SQL under cursor: runs the engine's EXPLAIN and flags full table scans / missing indexes.
• Suggest Index for SQL under cursor (command palette) — runs EXPLAIN, and if it finds a problem, asks the AI agent to propose a Laravel migration that adds the missing index.

Schema-aware completion suggests table and column names as you type SQL.

¶The SQL console

The query box in the results panel is a full SQL editor: syntax highlighting, schema-aware completion (tables, columns, keywords) as you type, and a draggable handle below it to resize.

• ⌘↵ runs the selection, or the statement under the cursor; ⌘⇧↵ (or Run) runs the whole console. Each statement gets its own result tab — pin one (★) to keep it across runs.
• :param placeholders prompt for values before running (remembered).
• ○ History opens a searchable log of every query you've run in this project; click one to load it back into the console.
• Long queries show a ✕ Cancel while running.
• Export the result as CSV / JSON / SQL, or copy it as TSV / Markdown.

¶Environments & safety

Each connection is labelled local, staging or production (green / amber / red), shown as a dot + badge in the list and on the active result. The label is guessed from the host and name and drives the safety rails:

• Destructive statements (DROP, TRUNCATE, DELETE/UPDATE without a WHERE) and any write to a non-local database open a confirmation dialog that lists the exact statements; non-local writes require an explicit acknowledgement.
• Transactional editing: cell edits and row deletes are staged (edited cells turn amber, deleted rows red). A N pending changes bar offers Submit (all in one transaction, via the confirmation dialog) and Revert. A Log button shows the session's writes with an Undo where a reverse statement can be generated.
• Snapshots: the ⤓ action on a local connection dumps the database (SQLite copy / mysqldump / pg_dump) into ~/.config/e/snapshots/ — handy before a migration.

¶Structure, relationships & tooling

• The Structure tab lists columns and indexes; Copy DDL copies the CREATE TABLE, and + Migration scaffolds a Laravel migration for the table and opens it (instead of running DDL directly).
• The ⇄ button in the panel header shows schema relationships (every foreign key as table.column → ref_table.ref_column).
• The cell dialog can Follow FK → (jump to the referenced row) and show Related → rows (reverse foreign keys).
• Seed 10 (local tables) creates rows via the Eloquent factory through Tinker; ⬆ CSV imports a CSV into the table.
• Search all data… in the panel header scans every table's text columns for a value and shows one result tab per matching table.
• Row counts appear next to each table in the tree.

¶Connection actions

Hover a connection for its actions: ⌗ new query, ⟳ refresh tables, ⤓ snapshot, ⏏ disconnect, ✎ edit, ✕ remove.