The desktop Firebase GUI for everyday development
Less Firebase busywork. More time to build.
Browse and edit Firestore, find users and follow failed requests in focused function logs. Keep it all open in tabs and pick up exactly where you left off.
Start free. No account. No expiry.
This video plays from YouTube. Google then receives your IP address, may store cookies or similar data in your browser and may process data in the USA.
Stop reopening what you were just working on.
Checking an order often means checking a customer and a function log, too. Keep each view in its own tab and switch back to your loaded work. Check the customer, investigate the request and return to the order without starting over.
- Keep your place in tabs
- Switch between loaded data, users and logs. Duplicate a view or reopen a closed tab. Connection colors keep the project recognizable.
- Go straight to the document
- Paste a document path into the address bar. Jump to collections, pinned paths and saved queries from the command palette.
- Build the query once
- Save a useful query and rerun it. Return to query history, or copy the query as SDK code for your application.
- Make repeat checks a board
- Keep query tables, charts and notes together. Change a parameter and refresh.See Boards in action →
One workspace for your daily Firebase tools
- Firestore
- Realtime Database
- Authentication
- Storage
- Cloud Functions
- Boards
- Security Rules
- Emulator
Boards for your Firebase data
Your daily checks, in one reusable dashboard.
Checking revenue, open orders and payment status should not mean rebuilding the same queries every morning. Bring totals, charts, query tables and notes together in a board. Open your saved view and refresh the data when you need it.
- See the numbers and the documents
- Keep summary metrics beside the matching records. Track order totals, spot a change in the chart and inspect the rows behind it.
- Reuse the view for a different question
- Add parameters for a customer, status or date. Change the inputs and rerun the cards instead of building another dashboard.
· boards · orders overview
One board for the daily order check. Revenue, order count, pending payments, daily orders, status breakdown and recent records. Real app, example data.
Find the error without losing the context.
A wall of log entries should not stand between you and the failed request. Filter by function, severity, text and time. Open one invocation to see the messages leading up to its error, then return to the document or user in the next tab.
- Hide request and system entries when you only need application messages.
- Expand structured payloads instead of reading flattened objects.
- Follow new entries with live polling while the view is open.
Invocation grouping requires an execution ID in your logs.
· function logs
An order confirmation failed among other function activity. Time, severity, function and message line up in one timeline. Real app, example data.
Data tools
Routine data jobs, without the one-off script.
Need a CSV for a colleague, test data in staging or the same fix across many documents? Export, copy and bulk-edit from the app. Collection copies and bulk edits require a dry run in the dialog, so you can review the planned changes before writing.
- Get readable files out of Firestore
- Export a collection or query result as JSON, JSONL, CSV or FS-JSON. Firestore export guide →
- Prepare staging with real project data
- Choose the source, destination and write mode, then inspect the copy's dry run. Copy data between projects →
- Apply the same change across documents
- Select matching documents, define the update and review its effect before applying it. Explore bulk edits and data tools →
· data between projects · products · development -> staging
Copying the product catalog from development to staging. The dry run found 8 documents to write and 0 to skip. Nothing is written until you press “Write 8 documents”.
Production safeguards
Fix the data. Double-check the target.
Working quickly should still leave room to catch the wrong project. Environment labels keep development, staging and production distinct. New production connections start write-protected, so you can inspect data before enabling changes.
- Confirm before destructive changes
- Deletes, bulk changes and Rules deploys in production ask you to type the project ID.
- Follow the change back to its source
- Filter the local audit log by action, actor and result. App, agent and built-in chat entries are labeled separately.
- Recover supported edits
- Restore supported Firestore and Realtime Database edits from saved local undo snapshots. The safety guide explains coverage and local data settings.
· guardrail · production · typed project ID
The project ID, halfway typed. fernway-pr is not fernway-prod, so Confirm stays disabled.
Agent access
Let AI help. Keep control of Firebase.
Connect Claude Code, Cursor and other MCP clients with access you choose per project. Grant read and write rights separately for each client and connection. New clients start with no access. Writes outside development wait for your approval, and your Google credentials stay in the app’s Rust core.
82 tools, 38 of them write. Agent writes are logged as mcp:<client>.
· approval · production · firestore_write
An agent wants to write to production. Two updates in one commit, shown before they run. No answer within 180 s counts as Deny.
Make your next Firebase task easier.
Download for Windows, macOS or Linux. Connect your project and start with Free. No account required.
Try every Pro feature for 14 days in the app. No credit card required.