> ## Documentation Index > Fetch the complete documentation index at: https://docs.erna.ai/llms.txt > Use this file to discover all available pages before exploring further. # Create a reusable pipeline configuration > Create a reusable pipeline configuration for a playbook. Do not create a new pipeline for every run. Create once, wait until it is ready, then execute it repeatedly with playbook-specific input. Lifecycle: Playbook -> Pipeline -> Revision -> Run. - Playbook defines the task type and run input/output schema. - Pipeline stores reusable instructions, business rules, and connector configuration. - Revision is the built executable version of the pipeline. - Run executes a ready revision with one-off input. Scheduling example, create once: { "name": "Cafe rota", "playbookId": "scheduling", "instructions": "Business rules, staff context, constraints, preferences, and recurring scheduling rules.", "connectors": [] } Then run many times with POST /v1/workspaces/{workspaceId}/pipelines/{pipelineId}/runs and input.dateRange. ## OpenAPI ````yaml /openapi.json post /v1/workspaces/{workspaceId}/pipelines openapi: 3.1.0 info: title: Erna API version: 1.0.0 description: >- AI business operations platform API. Erna uses a reusable pipeline lifecycle: Playbook -> Pipeline -> Revision -> Run. - A playbook defines the repeatable task type, generation behavior, and run input/output contract. - A pipeline stores reusable instructions and configuration for one playbook. Do not create a new pipeline for every execution. - A revision is the built executable version of a pipeline definition. - A run executes the latest ready revision with playbook-specific input. For scheduling, create one pipeline with the reusable business rules, staff context, and constraints. Then run that pipeline repeatedly with date-specific input such as { "input": { "dateRange": { "start": "2026-05-11", "end": "2026-05-17" } } }. servers: - url: https://api.erna.ai description: Production security: - bearerAuth: [] paths: /v1/workspaces/{workspaceId}/pipelines: post: tags: - Pipelines summary: Create a reusable pipeline configuration description: >- Create a reusable pipeline configuration for a playbook. Do not create a new pipeline for every run. Create once, wait until it is ready, then execute it repeatedly with playbook-specific input. Lifecycle: Playbook -> Pipeline -> Revision -> Run. - Playbook defines the task type and run input/output schema. - Pipeline stores reusable instructions, business rules, and connector configuration. - Revision is the built executable version of the pipeline. - Run executes a ready revision with one-off input. Scheduling example, create once: { "name": "Cafe rota", "playbookId": "scheduling", "instructions": "Business rules, staff context, constraints, preferences, and recurring scheduling rules.", "connectors": [] } Then run many times with POST /v1/workspaces/{workspaceId}/pipelines/{pipelineId}/runs and input.dateRange. operationId: createPipeline parameters: - schema: type: string example: ws_123 required: true name: workspaceId in: path - schema: type: string minLength: 1 example: unique-idempotency-key required: true name: Idempotency-Key in: header requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/CreatePipelineRequest' responses: '202': description: Pipeline accepted content: application/json: schema: $ref: '#/components/schemas/PipelineResponse' '400': description: Invalid request content: application/json: schema: type: object properties: error: type: string connector: type: string flow: type: string providerError: type: string details: type: array items: type: object properties: path: type: string message: type: string required: - path - message message: type: string reason: type: string runId: type: string requiredActions: type: array items: type: object properties: id: type: string type: type: string enum: - connect_service connector: type: string href: type: string request: type: object properties: {} default: {} additionalProperties: {} required: - id - type - connector - href additionalProperties: false billing: type: object properties: checkoutPath: type: string required: - checkoutPath required: - error additionalProperties: false '403': description: Missing write access content: application/json: schema: type: object properties: error: type: string connector: type: string flow: type: string providerError: type: string details: type: array items: type: object properties: path: type: string message: type: string required: - path - message message: type: string reason: type: string runId: type: string requiredActions: type: array items: type: object properties: id: type: string type: type: string enum: - connect_service connector: type: string href: type: string request: type: object properties: {} default: {} additionalProperties: {} required: - id - type - connector - href additionalProperties: false billing: type: object properties: checkoutPath: type: string required: - checkoutPath required: - error additionalProperties: false '404': description: Workspace not found content: application/json: schema: type: object properties: error: type: string connector: type: string flow: type: string providerError: type: string details: type: array items: type: object properties: path: type: string message: type: string required: - path - message message: type: string reason: type: string runId: type: string requiredActions: type: array items: type: object properties: id: type: string type: type: string enum: - connect_service connector: type: string href: type: string request: type: object properties: {} default: {} additionalProperties: {} required: - id - type - connector - href additionalProperties: false billing: type: object properties: checkoutPath: type: string required: - checkoutPath required: - error additionalProperties: false security: - bearerAuth: [] components: schemas: CreatePipelineRequest: type: object properties: name: type: string minLength: 1 description: Human-readable name for this reusable pipeline. playbookId: type: string enum: - scheduling - reporting description: >- Playbook that defines the reusable pipeline type and run input contract. instructions: type: string minLength: 1 description: >- Reusable business rules and configuration used for every run of this pipeline. For scheduling, include staff context, constraints, preferences, and recurring rules here, not the one-off dateRange. connectors: type: array items: type: object properties: connector: type: string minLength: 1 description: Connector identifier selected from the connector catalog. connectionId: type: string minLength: 1 description: Stored connection id used by pipeline builds and runs. required: - connector - connectionId additionalProperties: false default: [] description: >- Stored external connections available during pipeline builds and runs. required: - name - playbookId - instructions additionalProperties: false PipelineResponse: type: object properties: pipelineId: type: string description: Reusable pipeline id. name: type: string description: Human-readable reusable pipeline name. playbookId: type: string enum: - scheduling - reporting description: Playbook that owns this pipeline and defines its run input contract. status: type: string enum: - building - ready - failed description: >- Current pipeline build status. Only ready pipelines can be run, but schemas may be visible before the build completes. latestRevisionId: type: - string - 'null' description: >- Latest ready executable revision id, or null while the pipeline is building. error: description: Public error details when status is failed. links: type: object properties: self: type: string runs: type: string revisions: type: string required: - self - runs - revisions additionalProperties: false definition: type: object properties: playbookId: type: string enum: - scheduling - reporting description: >- Playbook that defines the reusable pipeline type and run input contract. instructions: type: string description: >- Reusable business rules and configuration used for every run of this pipeline. Do not put one-off run inputs such as a scheduling date range here. connectors: type: array items: type: object properties: connector: type: string minLength: 1 description: Connector identifier selected from the connector catalog. connectionId: type: string minLength: 1 description: Stored connection id used by pipeline builds and runs. required: - connector - connectionId additionalProperties: false description: >- Stored external connections available during pipeline builds and runs. required: - playbookId - instructions - connectors additionalProperties: false inputJsonSchema: type: - object - 'null' properties: {} additionalProperties: {} description: >- JSON Schema for the input object accepted by pipeline runs. Fetch pipeline detail before running and send matching data under the run request input property. outputJsonSchema: type: - object - 'null' properties: {} additionalProperties: {} description: JSON Schema for run output once a run completes. latestRun: type: - object - 'null' properties: runId: type: string status: type: string enum: - running - complete - failed createdAt: type: string updatedAt: type: string links: type: object properties: self: type: string required: - self additionalProperties: false required: - runId - status - createdAt - updatedAt - links additionalProperties: false description: >- Latest pipeline run summary, if one exists. Fetch links.self to read full output or error details. required: - pipelineId - name - playbookId - status - latestRevisionId - links - definition - inputJsonSchema - outputJsonSchema - latestRun additionalProperties: false securitySchemes: bearerAuth: type: http scheme: bearer description: OAuth2 access token or API key ````