Browser Extension
Export and Copy as code
Export a Recording as a JSON file with a published schema, with or without document contents and never with tokens, and copy any captured Firestore query as code for the Client SDKs or the Admin SDKs.
Export a Recording
Export in the panel’s toolbar opens Export the Recording:
- Without document contents leaves out every field value. Paths, queries with their filter values, counts and sizes stay in the file: a document id or a filter value can still say something about a person, so read the file before you share it.
- Tokens and who signed in are never exported.
- Save JSON saves the file, named after the project and the time, such as
fuego-recording-acme-shop-20261008-091244.json.
A Recording export is handy to attach to an issue, to compare a page before and after a change, or to read with a script.
What the file holds
$schema,"format": "fuego-recording"and"version": 1.- When it was exported, and the extension’s version and browser.
- The page as its origin and path only: never its query string or its fragment, which may carry tokens.
- What the Recording saw of Firebase: the project ids, whether the page used the emulators, and the databases. Never the API key.
- The cost settings, the date of the prices and the summary: counts and the estimated cost.
- Every operation: when it happened (milliseconds since the Recording started), its kind, operation, database, path, its query as the JSON Fuego’s query builder uses (when it converts), its counts, outcome and sizes, and the documents it received or wrote, with their Firestore typed values as they came (unless you left the contents out).
- Auth calls as events only: what happened, never the uid, email, claims or token.
Tokens never reach the Recording in the first place: the recorder sees that a call carries one and keeps nothing of it.
{
"$schema": "https://fuegoapp.dev/schemas/recording-v1.json",
"format": "fuego-recording",
"version": 1,
"exportedAt": "2026-10-08T09:12:44.120Z",
"extension": { "version": "1.0.0", "browser": "chrome" },
"page": { "origin": "https://shop.example.com", "path": "/products" },
"firebase": {
"projectIds": ["acme-shop"],
"emulator": false,
"databases": ["(default)"]
},
"withDocumentContents": true,
"operations": [
{
"time": 412,
"kind": "firestore",
"operation": "listen",
"database": "(default)",
"path": "products",
"live": true,
"outcome": "ok",
"documents": [
{
"path": "products/mug",
"change": "added",
"fields": {
"name": { "stringValue": "Mug" },
"price": { "doubleValue": 12.5 }
}
}
]
}
]
}
The example leaves out the cost settings, the summary and some fields of the operation.
The schema
The format is described by a JSON Schema published at
https://fuegoapp.dev/schemas/recording-v1.json,
the address every file carries in $schema: editors that read JSON Schema
check and complete the file with it. New fields keep version 1, and readers
ignore fields they do not know; only a change that breaks readers would bring
a version 2, at a new address. A query in the file follows Fuego’s own query
schema, query-builder.schema.json.
Copy as code
The detail of a Firestore query or listener has Copy as code and a language menu:
| Client SDK | Admin SDK |
|---|---|
| TypeScript / JavaScript, Kotlin / Java, Swift | Node.js, Python, Go, Java, C#, PHP, Ruby |
The Code tab shows the code before you copy it, and the panel remembers the language you chose. For a listener, the code reads the query once (“The code reads the query once; the page listens to it.”). Code is generated for queries: a document read, a write or a pipeline query has none.
The same generators write the code in Fuego’s query builder, so a query copied here reads like one written there. To keep working on the query itself, Open in Fuego opens it in Fuego’s query builder: see Opening things in Fuego.
Related
- Recording and cost
- Querying Data: the query builder in Fuego
- Permissions, privacy and troubleshooting
