Developers & Integrations

Put Grain Checks Into The Data Development Workflow

A full-screen Diff Terminal reviews every pull request like a person would — then folds the reasoning into collapsible Check Strips: Model Grain, Join Cardinality, Metric Impact, Dashboard Usage, Result. Block high-risk grain changes directly in GitHub or your CI/CD pipeline before merge.

Open API DocsRequest Integration
--- a/models/fct_billing_events.sql
+++ b/models/fct_billing_events.sql
@@ -12,6 +12,8 @@ select
   c.customer_id,
   i.amount
 from customers c
+join contacts x
+  on x.customer_id = c.customer_id
 join invoices i
   on i.customer_id = c.customer_id
Model GrainPASS

fct_billing_events resolves to one row per invoice

Join CardinalityFAIL

new contacts join is 1:N — fanout introduced

Metric ImpactWARN

3 metrics reference this model: Active Customers, ARPC, NRR

Dashboard UsageWARN

2 dashboards render an affected metric

ResultFAIL

Blocked — fanout must be resolved or the join approved

Run Grain Check on any GitHub PR or CI job — high-risk Result strips can block the merge until the join is approved.

Focus Credentials

Scoped test keys for CI and local runs

Generate a key scoped to one project — grain checks run in CI without touching production credentials.

POST /v1/checks/run
{ "project_id": "p_northwind", "model": "fct_billing_events", "pr": 482 }
Integrations

Every warehouse and BI tool in one grain layer

dbtSnowflakeBigQueryPostgreSQLLookerMetabaseTableau
API Documentation

Authentication · interfaces · webhooks

Authentication

API keys & OAuth

Bearer gl_live_…. Generate Test Key issues a key scoped to sample data only.

Interfaces

Core endpoints

  • POST /v1/checks/run
  • GET /v1/models/{id}/grain
  • GET /v1/metrics/{id}/contract
  • POST /v1/joins/measure
Webhooks

Outbound

check.blocked, metric.impacted, change.echoed — verify X-GrainLux-Signature.

A Correct Query Can Still Produce The Wrong Metric.

Illuminate The True Grain Of Every Metric.

GrainLux is the AI Semantic Data Reliability Layer between your warehouse and every model, metric, and dashboard built on top of it — grain-aware, metric-native, explainable, and deterministic.

Inspect A MetricBook A Demo