FlowRunner
PricingContact
Theme
Start Free

Kintone

Database

Treat Kintone apps as a live database for your agents. They query records with conditions and pagination, create or update up to 100 records at once, post comments with mentions to record timelines, and read app schemas to build valid payloads dynamically.

10 actions API key available
Kintone website ↗ Platform Documentation ↗ Capability data verified 2026-07-27
A new order lands from the storefront
Get Form Fields reads the orders app schema so the payload maps to current field codes
Query Records looks for an existing record with the same order number
Add Record writes the order with customer, amount, and status
The flow flags orders above the team's approval threshold
The fulfillment lead reviews flagged orders and picks standard or expedited handling
Update Record sets the decision and Add Record Comment mentions the packing team

What This Integration Enables

Kintone's quiet achievement is that operations teams built their own systems of record with it: the order tracker, the vendor list, the inspection log, each one an app someone assembled without engineering. The gap is that those apps still get fed by hand. FlowRunner closes it by treating every Kintone app as a live database an agent can query, write, and annotate, and the connector's schema discovery is what makes that durable: an agent calls Get Form Fields first, learns the current field codes and types, and builds valid payloads against the app as it exists today, not as it existed when someone wrote a script. - Query records with real conditions, ordering, and pagination to feed reports and downstream steps - Create and update records one at a time, or up to 100 per request with the bulk actions - Enforce optimistic locking on updates so concurrent edits fail loudly instead of overwriting silently - Post comments with @mentions to a record's timeline so notifications land where the work lives - Discover apps and schemas dynamically with List Apps and Get Form Fields

Without FlowRunner

Apps updated by hand The Kintone apps that replaced the spreadsheets still depend on someone retyping data from other systems
Integrations break on schema changes A renamed field silently kills the sync script, and nobody notices until the numbers disagree
Notifications happen in hallways A record changes state and the person who needs to act finds out at the next meeting

With FlowRunner

Apps fed by every connected system Orders, payments, and bookings write themselves into the right app within seconds of the event
Agents read the schema before they write Get Form Fields tells the agent the current field codes and types, so payloads adapt instead of breaking
The timeline carries the workflow Comments with @mentions land on the record itself, notifying exactly the people who need to move

Use Case Scenarios

Orders flow in, exceptions flow to people

A [Shopify](/integrations/shopify) order fires the flow. Query Records confirms the order is new, Add Record writes it into the orders app, and for orders above the threshold, Add Record Comment mentions the fulfillment lead with the details. When a [Stripe](/integrations/stripe) payment confirms later, the flow finds the matching record with Query Records and marks it paid with Update Record, passing the revision so a simultaneous manual edit fails the update instead of being overwritten.

The bulk import that respects the schema

Every morning, the flow aggregates yesterday's qualified leads from [Typeform](/integrations/typeform) and the booking calendar. Get Form Fields pulls the lead app's current schema, the agent maps each lead's fields to the live field codes, and Add Records writes them in batches of up to 100. When someone renames a dropdown or adds a required field, the next run sees the change in the schema and adapts the payload, instead of failing at 2 am with a cryptic field error.

The weekly report that assembles itself

On Friday afternoon, Query Records pulls every record in Status = "Open" ordered by age, with the total count included. The flow writes the snapshot to [Google Sheets](/integrations/google-sheets), posts a summary to [Slack](/integrations/slack) with the oldest items called out, and drops Add Record Comment mentions on the records that have sat untouched past the SLA, tagging their owners directly on the timeline where they will see it Monday morning.

Human-in-Loop Highlight

Delete Records removes up to 100 records per request and cannot be undone. In an app that has served as the team's system of record for years, that is potentially a hundred order histories, inspection results, or vendor negotiations gone in one call. So a FlowRunner cleanup agent never decides what dies. A quarterly sweep queries for records that look abandoned: closed status, no timeline activity, no linked records. It posts the candidate list with counts and a sample, and asks the app owner to confirm. The owner strikes two records that look dormant but are evidence in an open dispute, approves the rest, and only then does Delete Records run. The query, the filtering, and the batching are machine work. The judgment about which history the team can afford to lose is not.

Agent processes routinely
Detects exception requiring judgment
Clear match Continues automatically
Ambiguous Routes to human via preferred channel
Human decides
Agent resumes with decision

Agent Capabilities

10 actions

Records

4
  • Get Record Retrieves a single record by ID, with every field code mapped to its type and value. The full-context read before a decision or an update.
  • Query Records Retrieves multiple records using Kintone query syntax: field conditions, ordering, and pagination, with an optional total count. A single request returns up to 500 records. The read behind every report and sweep on this page.
  • Add Record Creates a single record, mapping field codes to value wrappers. Calculated, lookup-mapped, status, and category fields are set by Kintone and must be omitted. Returns the new record's ID and revision.
  • Update Record Updates a single record by ID, changing only the supplied fields. Pass a revision to enforce optimistic locking, so an update against a stale copy fails instead of overwriting someone's edit.

Bulk Operations

3
  • Add Records Creates up to 100 records in one request from an array of record objects, returning IDs and revisions in order. The workhorse for imports and aggregation flows.
  • Update Records Updates up to 100 records in one request, each entry pairing a record ID with the fields to change. Used for batch status moves and enrichment passes.
  • Delete Records Permanently deletes up to 100 records by ID in a single request. This cannot be undone, which is exactly why this page's human gate exists for it.

Collaboration

1
  • Add Record Comment Posts a comment to a record's timeline, optionally mentioning users, groups, or organizations so they are notified. The step that turns a workflow event into a notification on the record itself.

Schema and Apps

2
  • Get Form Fields Retrieves an app's field definitions: code, label, type, required flags, and dropdown options. The discovery step that lets agents build valid payloads against the app as it exists today.
  • List Apps Lists the apps available to the authenticated user with ID, name, description, and space, with name filtering and pagination. Used to resolve app IDs before any record operation.

Frequently Asked Questions

What can FlowRunner do with Kintone?

FlowRunner agents can run Get Record, Query Records, and Add Record in Kintone, plus 7 more actions.

Does connecting Kintone to FlowRunner require OAuth?

No. Kintone connects to FlowRunner with an API key, no OAuth flow required.

Can Kintone trigger a FlowRunner workflow automatically?

Kintone doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with Kintone

$100 in credits. No card required. Connect in minutes.