Create AskFred Thread

Create a new AskFred conversation thread with a question about your meetings

Overview

The createAskFredThread mutation allows you to start a new AskFred conversation by asking questions about your meeting transcripts. You can ask about a specific meeting or search across multiple meetings using filters. AskFred uses AI to analyze your meeting data and provide intelligent, context-aware answers.

Arguments

input · CreateAskFredThreadInput

Show child attributes

query · String · required

Your question or query about the meeting(s). Maximum 2000 characters.

transcript_id · String

ID of a specific transcript/meeting to query. If provided, the question will be answered based only on this meeting.

filters · AskFredMeetingFiltersInput

Filters to search across multiple meetings. Only used if transcript_id is not provided.

Show child attributes

start_time · String

Filter meetings from this date/time (ISO 8601 format)

end_time · String

Filter meetings until this date/time (ISO 8601 format)

channel_ids · [String]

Filter by specific channel/integration IDs

organizers · [String]

Filter by meeting organizer email addresses

participants · [String]

Filter by participant email addresses

transcript_ids · [String]

Filter by specific transcript IDs

response_language · LanguageCode

Language code for the response (e.g., ‘en’ for English, ‘es’ for Spanish). See Language Codes for full list.

format_mode · FormatMode

Response format: ‘markdown’ for rich formatting or ‘plaintext’ for simple text

Returns

Returns an AskFredResponse object containing the generated message with the answer to your query.

Usage Example

Query a Specific Meeting

mutation CreateThreadForMeeting($input: CreateAskFredThreadInput!) {
  createAskFredThread(input: $input) {
    message {
      id
      thread_id
      query
      answer
      suggested_queries
      status
      created_at
    }
  }
}

Query Across Multiple Meetings

mutation CreateThreadWithFilters($input: CreateAskFredThreadInput!) {
  createAskFredThread(input: $input) {
    message {
      id
      thread_id
      query
      answer
      suggested_queries
      status
    }
  }
}

curl

curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your_api_key" \
  -d '{
    "query": "mutation CreateThread($input: CreateAskFredThreadInput!) { createAskFredThread(input: $input) { message { id thread_id query answer suggested_queries status } } }",
    "variables": {
      "input": {
        "query": "What were the main decisions made in the product planning meeting?",
        "transcript_id": "transcript_xyz789",
        "response_language": "en",
        "format_mode": "markdown"
      }
    }
  }' \
  https://api.fireflies.ai/graphql

Response

{
  "data": {
    "createAskFredThread": {
      "message": {
        "id": "msg_001",
        "thread_id": "thread_new123",
        "query": "What were the main decisions made in the product planning meeting?",
        "answer": "Based on the product planning meeting transcript, here are the main decisions made:\n\n## Product Roadmap\n- **Q4 Focus**: Prioritize mobile app improvements and API v2 development\n- **Feature Freeze**: No new features after November 15th to focus on stability\n\n## Resource Allocation\n- **Team Expansion**: Approved hiring 2 senior engineers and 1 UX designer\n- **Budget**: Allocated $150K for Q4 development initiatives\n\n## Timeline\n- **Beta Release**: Scheduled for December 1st\n- **Public Launch**: Targeted for January 15th\n\n## Strategic Partnerships\n- **Integration Partners**: Approved partnerships with Slack and Microsoft Teams\n- **API Access**: Will provide early access to 5 strategic partners",
        "suggested_queries": [
          "What specific mobile app improvements were discussed?",
          "Who will be responsible for the API v2 development?",
          "What were the concerns raised about the timeline?"
        ],
        "status": "completed",
        "created_at": "2024-03-15T10:30:00Z"
      }
    }
  }
}

Error Codes

invalid_arguments

Validation error occurred. Common causes:

• Query exceeds 2000 character limit or is empty
• Both transcript_id and filters are provided (only one is allowed)

object_not_found

The specified transcript_id does not exist or you don’t have access to it

require_ai_credits

You don’t have access to AI credits. Upgrade your plan to use AskFred.

invalid_language_code

The provided response_language code is not supported. See Language Codes for valid options.

Related

Continue Thread

Add follow-up questions to continue the conversation

List Threads

View all your AskFred conversation threads