> ## Documentation Index > Fetch the complete documentation index at: https://api-docs.goextrovert.com/llms.txt > Use this file to discover all available pages before exploring further. # Generate replies > Generates a fresh Extrovert reply for each listed conversation, exactly as Generate does for one. Each conversation is generated on its own and gets its own result in request order — the reply with its dmId and text, or the reason it failed; a failed conversation never hides or undoes the others. Conversations of the same account are generated one after another, so a batch for one account takes longer than one spread over several accounts. Batch generation never takes instructions and never improves; nothing is sent. Up to 10 conversations per call. ## OpenAPI ````yaml https://api.goextrovert.com/client/v2/docs-json post /client/v2/dm-conversations/generate-replies openapi: 3.1.0 info: title: Extrovert Client API description: >- Public API for Extrovert platform integration. Use your API key for authentication. version: '1.0' contact: {} servers: - url: https://api.goextrovert.com/ description: API Server security: [] tags: [] paths: /client/v2/dm-conversations/generate-replies: post: tags: - Direct Messages summary: Generate replies description: >- Generates a fresh Extrovert reply for each listed conversation, exactly as Generate does for one. Each conversation is generated on its own and gets its own result in request order — the reply with its dmId and text, or the reason it failed; a failed conversation never hides or undoes the others. Conversations of the same account are generated one after another, so a batch for one account takes longer than one spread over several accounts. Batch generation never takes instructions and never improves; nothing is sent. Up to 10 conversations per call. operationId: DmConversationsClientApiController_generateReplies parameters: [] requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/GenerateRepliesRequestDto' responses: '200': description: '' content: application/json: schema: $ref: '#/components/schemas/GenerateRepliesResponseDto' '401': description: Invalid or missing API key. content: application/json: schema: $ref: '#/components/schemas/ClientApiErrorResponseDto' security: - api-key: [] components: schemas: GenerateRepliesRequestDto: type: object properties: items: type: array minItems: 1 maxItems: 10 items: type: object properties: connectionId: type: string format: uuid description: Conversation (connection) to generate a reply for. campaignId: type: string description: >- One-call Voice source: an updatable campaign in this workspace belonging to the conversation owner. Does not reassign the prospect. Omit for implicit selection, which may differ from the recommendation. format: uuid required: - connectionId description: Conversations to generate replies for, 1 to 10. required: - items GenerateRepliesResponseDto: type: object properties: status: type: string const: success description: Response status data: type: array items: type: object properties: index: type: integer minimum: 0 maximum: 9007199254740991 description: Position of the element in the request list, starting at 0. id: description: >- Element ID. Present for every updated or deleted element, and for a created element once it was saved. type: string format: uuid item: description: The element as saved. Absent when the element failed. type: object properties: dmId: type: string format: uuid description: Exact reply ID required by Edit, Send, and Recover. examples: - 8d256aa6-9b12-49a7-93ee-bf41634b5b60 connectionId: type: string format: uuid description: Conversation this reply belongs to. examples: - 8d256aa6-9b12-49a7-93ee-bf41634b5b60 text: type: string minLength: 1 description: Effective reply text. source: type: string enum: - Extrovert-generated - Agent-authored - Edited description: >- Extrovert-generated when no authored override exists; Agent-authored when created without generation; Edited when custom text overrides a generated reply. state: type: string enum: - No draft - Suggested - Draft ready - Generating - Sending - Sent - History changed - Error - Paused - Skipped allowedActions: type: array items: type: string enum: - prepare-reply - generate-or-improve-reply - edit-reply - send-reply - recover-draft description: Actions valid for the current state. reason: anyOf: - type: string enum: - Relevant post - Follow-up - First outreach - Manual - type: 'null' description: >- Why Extrovert recommended this conversation, or null when no automatic reason applies. followUpNumber: anyOf: - type: integer minimum: 1 maximum: 9007199254740991 - type: 'null' description: >- Which follow-up of the unanswered series is due, counting from 1. Null outside a follow-up series. examples: - 1 - 3 relevantPost: anyOf: - type: object properties: summary: anyOf: - type: string - type: 'null' description: >- Stored post summary, or null when none was persisted. linkedInUrl: type: string format: uri description: LinkedIn URL of the post the reply was written on. publishedAt: type: string description: When the post was published on LinkedIn. examples: - '2025-06-30T14:02:07.456Z' - '2025-07-03T08:30:00.000Z' required: - summary - linkedInUrl - publishedAt - type: 'null' description: >- Post the reply text was written on, or null when it was not written on a post. error: anyOf: - type: object properties: code: type: string description: >- Stable public error code suitable for branching. It is not a raw provider or backend status. message: type: string description: >- Safe explanation of what failed and what the caller should do next. retryable: type: boolean description: >- Whether retrying may succeed after the stated condition is resolved. occurredAt: anyOf: - type: string description: Date and time in ISO 8601 format examples: - '2025-06-30T14:02:07.456Z' - '2025-07-03T08:30:00.000Z' - type: 'null' description: >- When this failure was recorded, or null when no reliable timestamp exists. required: - code - message - retryable - occurredAt - type: 'null' description: >- Current send/workflow failure, or null when no failure is active. createdAt: type: string description: Date and time in ISO 8601 format examples: - '2025-06-30T14:02:07.456Z' - '2025-07-03T08:30:00.000Z' updatedAt: type: string description: Date and time in ISO 8601 format examples: - '2025-06-30T14:02:07.456Z' - '2025-07-03T08:30:00.000Z' sentAt: anyOf: - type: string description: Date and time in ISO 8601 format examples: - '2025-06-30T14:02:07.456Z' - '2025-07-03T08:30:00.000Z' - type: 'null' description: LinkedIn send time, or null before successful delivery. required: - dmId - connectionId - text - source - state - allowedActions - reason - followUpNumber - relevantPost - error - createdAt - updatedAt - sentAt error: description: >- Why this element was not applied. Absent when the element succeeded; the other elements are unaffected. type: object properties: code: type: string description: >- Status name of the failure, e.g. `NOT_FOUND`, `PERMISSION_DENIED`, `INVALID_ARGUMENT`, `INTERNAL`. message: type: string description: Human-readable reason this element was not applied. required: - code - message required: - index description: >- One result per requested conversation, in request order: the generated reply, or the reason it was not generated message: type: string description: Success message examples: - Operation completed successfully required: - status - data - message ClientApiErrorResponseDto: type: object properties: status: type: string description: Response status example: error enum: - error data: type: object description: Response data example: null nullable: true message: type: string description: Error message example: An error occurred while processing the request required: - status - data securitySchemes: api-key: type: apiKey in: header name: X-API-Key description: API Key for authentication ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.