Databox
Analytics & DataPush live KPIs into Databox dashboards from any workflow. Agents send metric values through the Push API, list metric keys, and audit or purge recent pushes to keep Custom data sources clean.
What This Integration Enables
A dashboard is only as trustworthy as the pipeline feeding it, and most KPI pipelines are a person with a CSV and good intentions. Databox's Push API turns the dashboard into a destination any workflow can write to, and FlowRunner turns that write into a governed step: computed from source systems, sanity-checked before it ships, verified after it lands, and correctable with an audit trail when something goes wrong. The number on the wall stops being a decoration and starts being a record. - Push KPIs from any workflow the moment the underlying event happens - Attach dates, units, and dimensions so one metric tells the story by channel or segment - Verify pushes landed and inspect exact payloads when a chart looks wrong - Discover what metric keys a data source already carries before adding more - Reset a polluted data source cleanly instead of layering corrections on bad data
Without FlowRunner
With FlowRunner
Use Case Scenarios
Revenue on the dashboard before the standup
Every morning the agent pulls yesterday's charges from [Stripe](/integrations/stripe), computes net revenue and new subscription count, and calls Push Metrics with yesterday's date and a currency unit. Get Last Pushes verifies both metrics landed without errors, and a confirmation posts to [Slack](/integrations/slack). By the time the team sits down, the dashboard already agrees with the billing system, and nobody has opened a spreadsheet.
The order counter that ticks in real time
Each new order in [Shopify](/integrations/shopify) fires a workflow that calls Push Metrics to increment the orders KPI and add the order value to revenue, dimensioned by sales channel. During a launch, the wall dashboard moves with the storefront instead of trailing it by a day. When the launch retro asks which channel actually carried the spike, the dimension attributes already hold the answer.
Recovering from a poisoned metric
A misconfigured upstream report pushed inflated conversion numbers for a week. The agent calls Get Metric Pushes for the affected key and lays out exactly which values arrived on which dates. The analytics owner reviews the damage and approves the reset. Only then does the agent run Purge Last Pushes to clear the source, recompute the correct series from [Google Analytics](/integrations/google-analytics), and re-send it with Push Raw Data, using full control of dates and dimensions to rebuild the history. The dashboard heals with a documented before and after instead of a quiet overwrite.
Human-in-Loop Highlight
Purge Last Pushes deletes all recent pushes for a data source, and with them the metric values they carried. It cannot be undone, and it is aimed at the exact numbers executives glance at between meetings. A reset is sometimes the right call, but it is never the agent's call. The flow presents the evidence: "Metric conversion_rate received 7 pushes between July 20 and 27 averaging 3x the trailing norm. Purging clears 214 pushes across this source. Purge and rebuild?" The analytics owner answers, and only then does the purge run, followed immediately by the clean re-send. The [human-in-the-loop](/concepts/human-in-the-loop) gate is what makes a destructive reset an act of maintenance rather than an incident.
Agent Capabilities
7 actionsPushing Metrics
2- Push Metrics Sends one or more metric values using a friendly per-metric structure with name, numeric value, and optional date, unit, and dimension attributes. FlowRunner assembles the Databox push payload for you, and hundreds of metrics fit in a single call.
- Push Raw Data Sends a raw push payload with full control of the record shape, including the "$"-prefixed metric keys and dimension fields. The tool for precise backfills and advanced record structures.
Auditing
3- List Metric Keys Lists every metric key that has been pushed to the data source with its most recent push metadata. The inventory step before adding metrics or debugging a dashboard.
- Get Last Pushes Retrieves the most recent pushes newest first, including payloads, HTTP responses, and any errors Databox reported. The verification step after a push and the first stop when a chart looks wrong.
- Get Metric Pushes Retrieves recent pushes for a single metric key, newest first. The forensic view when one KPI needs its exact history inspected.
Cleanup
2- Delete Push Permanently deletes a single push by its ID, removing its metric values from your data. Used to surgically remove one bad value, with the ID sourced from Get Last Pushes.
- Purge Last Pushes Deletes all recent pushes for the data source and the values they contain. The operation this page's human gate exists for.
Frequently Asked Questions
What can FlowRunner do with Databox?
FlowRunner agents can run Push Metrics, Push Raw Data, and List Metric Keys in Databox, plus 4 more actions.
Does connecting Databox to FlowRunner require OAuth?
No. Databox connects to FlowRunner with basic authentication (a username and password), no OAuth flow required.
Can Databox trigger a FlowRunner workflow automatically?
Databox doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.
Start building with Databox
$100 in credits. No card required. Connect in minutes.