Elnora Tasks

Elnora Tasks

Tasks are conversation threads with the Elnora AI Platform. Each task is a chat where you send messages and receive AI-generated protocol responses. Use tasks to generate, iterate on, and refine bioprotocols.

Organization Context

All list and create tools accept an optional org_id parameter (UUID). When provided, the operation targets that organization instead of the user's active org. The user must be a member of the target org.

Concepts

• Task: A conversation thread tied to a project. Has a title, status, and message history.
• Message: A single turn in the conversation. Each has a role (user or assistant), sequence number, optional attachments, and content.
• Protocol generation: Create a task, send a message describing what you need, and the assistant responds with a generated protocol. Iterate by sending follow-up messages.
• All IDs are UUIDs (e.g., bfdc6fbd-40ed-4042-9ea7-c79a5ec90085).

MCP Tools

elnora_list_tasks

List tasks, optionally filtered by project or status.

Returns paginated results:

{
  "items": [
    {
      "id": "<UUID>",
      "projectId": "<UUID>",
      "title": "...",
      "status": "active",
      "messageCount": 4,
      "lastMessageAt": "...",
      "createdAt": "..."
    }
  ],
  "page": 1,
  "pageSize": 25,
  "totalCount": 12,
  "totalPages": 1,
  "hasNextPage": false
}

elnora_get_task

Get full details for a single task.

Returns task detail including metadata, status, and associated project. Use this to inspect a task before reading its messages.

elnora_create_task

Create a new task (conversation thread).

Returns the created task object with its id. Use this ID for all subsequent operations (send messages, get messages, update, archive).

To start protocol generation immediately, provide initial_message. To create an empty task for later use, omit it.

elnora_send_message

Send a message to a task and receive the AI response. May take 30-120 seconds for complex protocol generation requests.

Returns the AI response. Use file_ids to reference uploaded files (templates, datasets) that should inform the protocol generation.

elnora_get_task_messages

Get message history for a task. Uses cursor-based pagination.

Returns:

{
  "items": [
    {
      "id": "<UUID>",
      "role": "user",
      "content": "Generate a PCR protocol for BRCA1 exon 11",
      "sequence": 1,
      "createdAt": "...",
      "attachments": []
    },
    {
      "id": "<UUID>",
      "role": "assistant",
      "content": "# PCR Protocol for BRCA1 Exon 11\n...",
      "metadata": "{\"status\":\"completed\"}",
      "sequence": 2,
      "createdAt": "...",
      "attachments": []
    }
  ],
  "nextCursor": null,
  "hasMore": false
}

Messages are ordered by sequence. If hasMore is true, pass nextCursor as cursor in the next call.

elnora_update_task

Update a task's title or status.

Must provide at least one of title or status.

elnora_archive_task

Archive (permanently delete) a task. This is destructive and cannot be undone. Always confirm with the user before calling.

Agent Workflow Recipes

Generate a protocol (full flow)

1. Find or create a project:
   • Call elnora_list_projects to get a project ID
   • Or call elnora_create_project if needed
2. Create a task with an initial prompt:
   • Call elnora_create_task with project_id, title, and initial_message
   • Save the returned task id
3. Read the AI response:
   • Call elnora_get_task_messages with the task ID
   • Look for the message with role: "assistant"
4. Iterate on the output:
   • Call elnora_send_message with refinement instructions
   • Read updated messages to see the new response

Quick protocol generation (shortcut)

For simple one-shot generation, use elnora_generate_protocol instead:

This creates a task, sends the description, waits for the response, and returns the result in one call. Takes 30-120 seconds. Use the full flow above when you need to iterate.

Check the latest assistant response

1. Call elnora_get_task_messages with limit: 2
2. Find the message where role is "assistant" -- this is the most recent AI output

Iterate on a protocol with file context

1. Get the file ID from elnora_list_files or elnora_search_files
2. Call elnora_send_message with the task ID, your refinement message, and file_ids containing the reference file
3. The AI will use the file content to inform its response

Pagination for long conversations

1. Call elnora_get_task_messages with limit: 50
2. If hasMore is true, call again with cursor set to nextCursor
3. Repeat until hasMore is false