restdb.io
Databaserestdb.io is a hosted NoSQL database whose REST API is generated from your schema, with MongoDB-style querying. Agents list, create, replace, update and delete documents, read and add child documents, query and manage media, read metadata and send email from the database.
What This Integration Enables
restdb.io is a backend, not a workspace. Its REST API is generated from the schema you define, it queries in MongoDB style, it can send its own email, and it can mint tokens for people who are not restdb.io users at all. That combination is why it tends to sit behind something public: a site, a portal, a form, a small product. The database is not where a team keeps its notes. It is where an application keeps its state.
Automating a backend calls for different reflexes than automating a workspace. Field names and collection names are case sensitive here, and a mismatch returns nothing rather than an error, which is why Get Database Metadata earns its place at the top of a flow. PUT and PATCH sit on the same document URL and do different things, so this connector names them Replace Document and Update Document rather than leaving the distinction in the documentation. And the bulk delete refuses an empty query outright, because an unbound template expression produces exactly the empty parameter that would otherwise clear a whole collection. FlowRunner's connectors are built and verified against each vendor's official API, and on restdb.io that verification is largely a matter of making the destructive shapes visible in the operation picker instead of in a footnote.
Without FlowRunner
With FlowRunner
Use Case Scenarios
-
A portal whose data stays coherent
A form on a site built in Webflow posts into a requests collection. The agent picks it up, calls Get Collection Metadata to confirm the schema, and runs List Documents with a query on the email address to decide whether this is a new contact or an existing one. Existing contacts get Update Document, which writes only what changed. Either way, Create Child Documents appends the submission to that contact's history sub collection, so the record carries its own timeline rather than accumulating parallel top-level rows. Send Email acknowledges the submission using the database's own mail service, without a separate provider in the path.
-
Reconciling application state against payments
On a schedule the agent runs List Documents with a query for subscriptions in an active state, pulls the matching charges from Stripe, and compares. Records that agree are left alone. Records that disagree get flagged with Update Document and posted to Slack with both sides shown. Nothing is corrected automatically, because a mismatch between what an application believes and what a payment processor recorded is a question about which one is wrong, and that is not a question a filter can answer.
-
A re-import that means it
When an authoritative export from another system has to overwrite a collection, Replace Document is the correct call and Update Document is not, because the point is that omitted fields should go. The agent writes each document in full, and the flow says so explicitly in its own step name. Create Documents with Skip Validation is available for the bulk load case, and the connector treats it as an opt-in that has to be declined properly rather than a flag that is truthy whenever it is present.
Human-in-Loop Highlight
The media archive is where this API stops being ordinary. GET /media/{id} requires no API key on restdb.io's side, which means a media URL is effectively public to anyone who has it, while the metadata routes around it are not. So Upload Media is not a storage operation, it is a publishing operation, and there is no undoing it: Delete Media removes the file, but it cannot recall a URL that has already been fetched or forwarded. A FlowRunner agent handling documents therefore uploads freely where the content is already public and stops where it is not: "This flow is about to call Upload Media on 34 files attached to claim records, including scanned identity documents. Once uploaded, each file is retrievable at its media URL by anyone holding that URL, with no key. Confirm these files are cleared for that, or route them to restricted storage instead?" The operations owner answers. The agent is perfectly capable of moving the bytes. What it cannot do is judge whether a document should exist at an address that is public to whoever receives it, and that judgment is exactly the kind FlowRunner is built to hand back.
Agent Capabilities
21 actionsDocuments
7- List Documents Lists documents from a collection using MongoDB style query syntax, with the modifiers that control paging, sorting, and field selection. A query returns at most 1,000 documents unless the maximum is raised.
- Get Document Returns one document by its ObjectID. Use List Documents with a query when the flow knows a business key rather than an ObjectID.
- Create Documents Creates one or more documents in a collection, taking a single object or an array. Skip Validation bypasses the collection's schema checks for a bulk load, which writes rows that would otherwise be rejected.
- Replace Document Replaces a document wholesale: the body becomes the document, and every field left out is removed. Reach for this only when the record is genuinely meant to be exactly what is being sent.
- Update Document Updates only the properties supplied and leaves every other field as it was. The safe edit and the default.
- Delete Document Permanently deletes one document by its ObjectID. There is no undo on this API.
- Delete Documents Deletes many documents at once, either by a list of ObjectIDs or by a query. An empty query would match the entire collection, so this connector refuses one outright.
Child Documents
3- List Child Documents Lists the documents in a sub collection of a parent document. A sub collection is a field of type child, so the name here is the field name on the parent.
- Get Child Document Returns one document from a sub collection of a parent, by its ObjectID.
- Create Child Documents Creates one or more documents inside a sub collection of a parent. The operation that gives a record its own timeline rather than scattering related rows across the top level.
Metadata
2- Get Database Metadata Returns the schema of the whole database: every collection with its fields and their types. Worth calling first, because collection and field names are case sensitive and a mismatch returns nothing rather than an error.
- Get Collection Metadata Returns one collection's fields, types, and settings, including which are lookups or child relations. This is the schema a document must match when validation is on.
Media
5- Upload Media Uploads a file into the database's media archive and returns its ObjectID. Reference that ObjectID from a document field to attach it, because uploading alone attaches the file to nothing.
- Download Media Downloads a file from the media archive at the size you choose. This route requires no API key on restdb.io's side, which is what makes a media URL effectively public.
- Get Media Metadata Returns a media object's original filename, size, MIME type, and upload date. Unlike the binary download, this route does require an API key.
- Query Media Searches the media archive with the same query syntax as the document routes, or fetches a specific list of media objects by identifier. Used to find files by size, type, or upload date.
- Delete Media Permanently deletes a file from the archive. Documents referencing its ObjectID keep the reference but it no longer resolves, so those fields need clearing too.
- Send Email Sends email through the database's own mail service, taking a single message or an array. It is how a restdb.io backed flow notifies people without wiring in a separate provider.
Authentication
3- Generate JWT Token Generates a token for an external user, signed with a secret configured in the database's global settings. How a restdb.io database authenticates people who are not restdb.io users.
- Get User Info Returns the email, display name, and image of the currently authenticated user. Used to confirm which identity a token resolves to.
- Logout User Logs a user out and permanently invalidates their login token. Logging out the token a connection uses will break every flow depending on it.
Frequently Asked Questions
What can FlowRunner do with restdb.io?
FlowRunner agents can run List Documents, Get Document, and Create Documents in restdb.io, plus 18 more actions.
Does connecting restdb.io to FlowRunner require OAuth?
No. restdb.io connects to FlowRunner with an API key, no OAuth flow required.
Can restdb.io trigger a FlowRunner workflow automatically?
restdb.io doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.
Other Database integrations
Start building with restdb.io
Free plan, no card required. Connect in minutes.