> ## Documentation Index > Fetch the complete documentation index at: https://docs.sagescreen.dev/llms.txt > Use this file to discover all available pages before exploring further. # Start full culture generation workflow > Starts a workflow run and asynchronously fires the full culture generation pipeline. Returns the workflow run payload immediately (~100ms). The actual generation happens in a background thread. The workflow type is determined by the `customizable` flag: - `true` or `1` → `sage` workflow (full AI pipeline) - `false` or `0` → `sage-nc` workflow (non-customizable, skips AI steps) ## OpenAPI ````yaml /openapi/private/specs/services/culture.yml post /culture/{council_sage_id}/save/full openapi: 3.1.0 info: title: Culture Service API description: > Manages the culture generation pipeline for council sages. Culture encompasses the tone, instructions, intro questions, and scoring configuration that define how a sage conducts screening interviews. ## Authentication All endpoints require authentication via one of two methods: - **API Key**: Pass the key in the `X-Auth` header (service-to-service calls) - **JWT Token**: Pass a Bearer token in the `Authorization` header (user-facing calls) The required permission level is noted per endpoint. ## Culture Pipeline 1. **Process** -- Collect all culture inputs (instructions, tone, duration, scoring) 2. **Save Full** -- Start a workflow run and asynchronously generate the culture 3. **Generate** -- Run the full generation pipeline (context analysis, tone crafting, intro questions, scoring) 4. **Save** -- Save individual culture step results to the Prompt Library 5. **End** -- Finalize the council sage after culture generation ## Business Rules - Culture data is stored in Redis during the generation pipeline - The `save/full` endpoint starts a workflow run and fires async generation - The `generate` endpoint validates payload via `CreateSageGenerate` schema - Customizable sages use the full pipeline; non-customizable sages skip AI tone/context steps - The `end` endpoint finalizes the council sage in the Prompt Library version: 1.0.0 contact: name: Platform Team servers: - url: '{protocol}://{host}:{port}' description: Culture Service variables: protocol: default: https enum: - http - https host: default: localhost port: default: '5000' security: [] tags: - name: Culture description: Culture CRUD, generation, and pipeline operations - name: Service description: Health checks and initialization paths: /culture/{council_sage_id}/save/full: post: tags: - Culture summary: Start full culture generation workflow description: > Starts a workflow run and asynchronously fires the full culture generation pipeline. Returns the workflow run payload immediately (~100ms). The actual generation happens in a background thread. The workflow type is determined by the `customizable` flag: - `true` or `1` → `sage` workflow (full AI pipeline) - `false` or `0` → `sage-nc` workflow (non-customizable, skips AI steps) operationId: saveCultureFull parameters: - $ref: '#/components/parameters/CouncilSageId' - $ref: '#/components/parameters/RequiredPermission_council_sage' requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/CreateSageGenerate' example: council_sage_id: c3d4e5f6-7890-4abc-def0-123456789abc context_id: a1b2c3d4-e5f6-4a7b-8c9d-0e1f2a3b4c5d name: Senior Software Engineer context: We are looking for a senior software engineer... role: Software Engineer role_level: senior instructions: Focus on problem-solving duration: 30 tone: - professional intro_question: Tell me about a challenging project intro_question_results: Evaluating depth scoring_low: 50 scoring_exclude: [] scoring_include: - 3 - 4 - 5 customizable: true responses: '200': description: Workflow run started, generation running asynchronously content: application/json: schema: type: object description: Workflow run data from the workflow service '400': description: Missing council sage ID or validation error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' examples: missingCouncilSageId: summary: Missing council_sage_id value: error: council_sage_id is required noJsonBody: summary: No JSON payload value: error: No JSON data provided '401': $ref: '#/components/responses/Unauthorized' '500': description: Error starting workflow or generation content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: >- Error starting culture generation: ConnectionError('Workflow service unavailable') security: - jwtAuth: [] components: parameters: CouncilSageId: name: council_sage_id in: path required: true description: UUID of the council sage schema: type: string format: uuid example: c3d4e5f6-7890-4abc-def0-123456789abc RequiredPermission_council_sage: name: X-Permission-Info in: header required: false description: >- **Required JWT permission: `council_sage`** (informational only -- enforced server-side) schema: type: string schemas: CreateSageGenerate: type: object required: - council_sage_id - context_id - name - role - role_level - duration - scoring_low properties: council_sage_id: type: string format: uuid description: UUID of the council sage context_id: type: string format: uuid description: UUID of the context name: type: string description: Display name for the sage context: type: string default: '' description: Job description text role: type: string description: Role title role_level: type: string description: Experience level instructions: type: string default: '' description: Interview instructions duration: type: integer description: Interview duration in minutes tone: type: array items: type: string default: [] description: Tone descriptors intro_question: type: string default: '' description: Opening question intro_question_results: type: string default: '' description: Evaluation criteria for intro question scoring_low: type: integer description: Low scoring threshold percentage scoring_exclude: type: array items: type: integer default: [] scoring_include: type: array items: type: integer default: [] customizable: type: boolean default: true description: >- Whether this is a customizable sage (true = full AI pipeline, false = non-customizable) language_code: type: - string - 'null' description: Language code for the interview ErrorResponse: type: object required: - error properties: error: type: string description: Human-readable error message responses: Unauthorized: description: Authentication failed content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' examples: noApiKey: summary: Missing API key value: error: Unauthorized - API key required invalidApiKey: summary: Invalid API key value: error: Unauthorized - Invalid API key noToken: summary: Missing JWT token value: error: Unauthorized - No valid token invalidToken: summary: Invalid or expired JWT value: error: Unauthorized3 insufficientPermissions: summary: Token lacks required permission value: error: Unauthorized4 securitySchemes: jwtAuth: type: http scheme: bearer bearerFormat: JWT description: User JWT token passed in Authorization header ````