Skip to content

title: "API overview" description: "The Minyu API provides programmatic access to all data defined in the Minyu data model." The API is GraphQL-based and is generated automatically from the co...

API overview

The Minyu API provides programmatic access to all data defined in the Minyu data model.
The API is GraphQL-based and is generated automatically from the configured schema, which means there is no separate API layer to design or maintain.

There is no API versioning. Any change to the data model—such as renaming tables, columns, or relations—is reflected in the API immediately and may break existing integrations if not coordinated carefully. All API requests are executed in the context of an authenticated user.

API scope

The API exposes all schema-defined entities and relations for both reading and writing.
It supports:

  • Single-row and collection queries
  • Relational traversal across entities
  • Create, update, and delete mutations
  • Nested mutations for editing relations

Each request is evaluated against live data and executed as a single, isolated operation.

Authentication

All API access requires authentication using Firebase-issued access tokens.
API requests are authenticated in the same way as interactive users, and all authorization rules are enforced accordingly.

In production environments, integrations typically use a dedicated API user to isolate system access from human user accounts. All requests are executed with the permissions of the authenticated user.

Schema mapping

The GraphQL schema is generated directly from the configured API names in the data model:

  • Tables define query and mutation entry points
  • Columns define scalar fields
  • Relations define nested fields and collections

Naming conventions follow the schema configuration exactly. Any schema change is reflected in the API immediately, without a deployment or migration step.

Queries

Queries provide read-only access to schema-defined data.

Key characteristics:

  • Read rules are enforced
  • Relational traversal is supported
  • Queries may return single rows or collections
  • All queries operate on live data

There are no artificial restrictions on query depth beyond what is enforced by read rules.

Mutations

Mutations provide write access to schema-defined data.

Key characteristics:

  • Write rules are enforced
  • Mutations are fully transactional per request
  • If any error occurs, the entire mutation is rolled back
  • Nested mutations are supported for editing relations

When mutating collections (relations with multiple items), each related item must specify an explicit _action to indicate whether it should be added, edited, removed, or deleted. This avoids ambiguity and prevents accidental replacement of large collections.

Rate limits

To protect platform stability, the API enforces request rate limits and execution time limits.

Limits apply per tenant and are enforced regardless of request type. Exceeding a limit causes the request to fail without applying partial changes. Rate limits are independent of usage-based pricing and apply to all API consumers.

Related concepts