Skip to main content
Glama

Superglue MCP

Official
by superglue-ai
overview.md6.16 kB
--- title: 'Overview' description: 'Overview of the superglue GraphQL API' --- The Core API provides GraphQL endpoints for managing API configurations, data extraction, transformations, and workflows. Main concepts: * **Workflows**: Chain multiple steps into a single execution (recommended) * **Integrations**: Manage integrations (e.g. Stripe, Hubspot) and their credentials ## Deprecated Operations **⚠️ The following operations are deprecated:** * **API Calls**: Individual `call` operations - use `executeWorkflow` instead * **Extractions**: Individual `extract` operations - use `executeWorkflow` instead * **Transformations**: Individual `transform` operations - use `executeWorkflow` instead * **Legacy Config Management**: `upsertApi`, `upsertExtract`, `upsertTransform` - use workflow-based operations instead These deprecated operations are still available but moved to the bottom of the documentation. Use workflows for better performance, reliability, and capabilities. ## Endpoint Use [`https://graphql.superglue.cloud`](https://graphql.superglue.cloud) or omit endpoint in the SDK. Self-hosted default port: 3000. ## Authentication All requests require a bearer token: ```http Authorization: Bearer YOUR_AUTH_TOKEN ``` ## Base Types ```graphql interface BaseConfig { id: ID! version: String createdAt: DateTime updatedAt: DateTime } union ConfigType = ApiConfig | ExtractConfig | TransformConfig | Workflow ``` ## Input Types ### ApiInput - id: ID! - urlHost: String! - urlPath: String - instruction: String! - queryParams: JSON - method: HttpMethod - headers: JSON - body: String - documentationUrl: String - responseSchema: JSONSchema - responseMapping: JSONata - authentication: AuthType - pagination: PaginationInput - dataPath: String - version: String ### ExtractInput - id: ID! - urlHost: String! - urlPath: String - queryParams: JSON - instruction: String! - method: HttpMethod - headers: JSON - body: String - documentationUrl: String - decompressionMethod: DecompressionMethod - fileType: FileType - authentication: AuthType - dataPath: String - version: String ### TransformInput - id: ID! - instruction: String! - responseSchema: JSONSchema! - responseMapping: JSONata - version: String ### WorkflowInput - id: String! - steps: [ExecutionStepInput!] - integrationIds: [ID!] - finalTransform: JSONata - inputSchema: JSONSchema - responseSchema: JSONSchema - version: String - instruction: String ### IntegrationInput - id: ID! - name: String - urlHost: String - urlPath: String - credentials: JSON - documentationUrl: String - documentation: String - documentationPending: Boolean (default: false) - specificInstructions: String ### RequestOptions Control how operations are executed with fine-grained options. ```json { "selfHealing": "ENABLED", // Default: ENABLED "cacheMode": "READONLY", // Deprecated, Default: READONLY "timeout": 300000, // Default: 300000ms (5 minutes) "retries": 10, // Default: 10 attempts "retryDelay": 0, // Default: 0ms (no delay) "webhookUrl": null, // Deprecated, Default: null (no webhooks) - this only works for calls, not for workflows. "testMode": false // Default: false - if this is true, superglue will validate the request after each execution. This is useful for building, testing and debugging. } ``` **Field Explanations:** - `selfHealing`: If it should auto-fix issues - `ENABLED`: Full auto-healing (recommended) - `TRANSFORM_ONLY`: Only fix data transformation issues - `REQUEST_ONLY`: Only fix API request issues - `DISABLED`: No auto-healing - `cacheMode`: Deprecated - If it should use the saved configuration and update it if self-healing is performed. this only works for calls, not for workflows. - `ENABLED`: Cache reads and writes - `READONLY`: Only read from cache, don't write (Default) - `WRITEONLY`: Only write to cache, don't read - `DISABLED`: No caching - `timeout`: Maximum time to wait (milliseconds) - `retries`: Number of retry attempts on failure - `retryDelay`: Delay between retries (milliseconds) - `webhookUrl`: POST endpoint for async notifications - this only works for calls, not for workflows. - `testMode`: If true, validate each request after execution. This is useful for building, testing and debugging. ### PaginationInput - type: PaginationType! - pageSize: String (default: "50") - cursorPath: String - stopCondition: String ### SystemInput - id: String! - urlHost: String! - urlPath: String - documentationUrl: String - documentation: String - credentials: JSON ## Enums ### HttpMethod GET, POST, PUT, DELETE, PATCH, HEAD, OPTIONS ### CacheMode ENABLED, READONLY, WRITEONLY, DISABLED ### FileType CSV, JSON, XML, AUTO ### AuthType NONE, OAUTH2, HEADER, QUERY_PARAM ### DecompressionMethod GZIP, DEFLATE, NONE, AUTO, ZIP ### PaginationType OFFSET_BASED, PAGE_BASED, CURSOR_BASED, DISABLED ### LogLevel DEBUG, INFO, WARN, ERROR ### SelfHealingMode ENABLED, TRANSFORM_ONLY, REQUEST_ONLY, DISABLED ### UpsertMode CREATE, UPDATE, UPSERT ## Common Parameters All execution operations (`executeWorkflow`) accept a `RequestOptions` object. ### Default Query Parameters Most list operations support: - `limit: Int` (default: 50) - `offset: Int` (default: 0) ## Error Handling All operations return: ```graphql { success: Boolean! error: String startedAt: DateTime! completedAt: DateTime! } ``` ## Retry Logic - API calls: up to 8 retries - Extractions: up to 5 retries - Each retry can generate a new config based on the previous error ## Webhooks If `webhookUrl` is set in options: - On success: POST `{success: true, data: result}` - On failure: POST `{success: false, error: message}` ## Workflows Workflows let you chain multiple steps (API, extract, transform) into a single execution. Each step can run in DIRECT mode or LOOP mode for batch processing. ## Integrations Integrations manage connections to third-party services, storing credentials and configuration needed for API calls. See also: - [Types Reference](types.md) - [Queries](queries.md) - [Mutations](mutations.md)

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/superglue-ai/superglue'

If you have feedback or need assistance with the MCP directory API, please join our Discord server