Google Cloud Firestore

Writes a new document at a specified path in Firestore, setting all fields in a single atomic write.

Upserts a document at a given path — creates it if absent, or merges the supplied fields if it already exists.

Permanently removes a document from Firestore by its full path, freeing storage and keeping the collection clean.

Fetches a single document by its full Firestore path, returning all fields at the moment of the read.

Retrieves every document in a named Firestore collection, returning the full set for downstream processing.

Executes a structured Firestore query with filters, ordering, and limits, returning only the matching document set.

Lists every top-level collection in the Firestore database, returning their names for discovery or introspection.

Patches specific fields on an existing Firestore document without overwriting unrelated fields.

Atomically commits up to 500 create, update, or delete operations in a single Firestore batch, succeeding or failing as one unit.

Executes a read-then-write sequence inside a Firestore transaction, guaranteeing consistency even under concurrent writes.

Opens a real-time listener on a Firestore document, delivering the current state and every subsequent change until the listener is closed.

Executes a server-side COUNT, SUM, or AVG aggregation over a Firestore collection, returning a single scalar result without downloading every document.

Returns the names of all subcollections beneath a given document, enabling dynamic traversal of nested Firestore data.

Queries across all Firestore subcollections sharing the same name, regardless of the parent document path.

Triggers a managed import or export of Firestore documents to or from a Google Cloud Storage bucket for backup, migration, or bulk load.

Fires when a new document is written to a watched Firestore collection for the first time.

Fires whenever one or more fields on an existing Firestore document change, giving the agent access to both the before and after state.

Fires when a document is removed from a watched Firestore collection, passing the final field values to the agent before they disappear.

Fires on any write to a watched document — whether it is being created for the first time or an existing document is being modified.

Fires when the document count in a watched Firestore collection crosses a configured high or low watermark.

Fires when a specific named field on a watched Firestore document transitions to a new value, allowing the agent to react to targeted data changes rather than any document update.

Fires on a recurring schedule to query all documents in a Firestore collection and process those matching a filter condition.