todoist_get_tasks

Instructions

Implementation Reference

• src/tasks.py:121-240 (handler)

  Main handler function implementing the todoist_get_tasks tool. Fetches tasks from Todoist API supporting filters by project, section, parent, label, or IDs, with pagination handling up to nmax tasks.

  def todoist_get_tasks(
      ctx: Context,
      project_id: Optional[str] = None,
      section_id: Optional[str] = None,
      parent_id: Optional[str] = None,
      label: Optional[str] = None,
      ids: Optional[list[str]] = None,
      nmax: Optional[int] = 100,
      limit: int = 200
  ) -> str:
      """Get a list of tasks from Todoist with basic filters
  
      This is a wrapper around the Todoist API's get_tasks method that handles pagination
      automatically. By default, it will fetch up to 100 matching tasks. Set nmax=None
      to fetch ALL matching tasks across multiple API calls.
  
      For natural language filtering (like 'today', 'overdue'), use todoist_filter_tasks instead.
  
      Examples:
          # Get up to 100 tasks (default)
          todoist_get_tasks(ctx)
  
          # Get all tasks in a project (up to 100 by default)
          todoist_get_tasks(ctx, project_id="12345")
  
          # Get first 500 tasks total
          todoist_get_tasks(ctx, nmax=500)
  
          # Get ALL tasks (unlimited)
          todoist_get_tasks(ctx, nmax=None)
  
          # Get specific tasks by ID
          todoist_get_tasks(ctx, ids=["task1", "task2", "task3"])
  
      Args:
          project_id: Filter tasks by project ID (optional)
          section_id: Filter tasks by section ID (optional)
          parent_id: Filter tasks by parent task ID (optional)
          label: Filter tasks by label name (optional)
          ids: A list of the IDs of the tasks to retrieve (optional)
          nmax: Maximum total number of tasks to return. Set to None for ALL matching tasks (default: 100)
          limit: Number of tasks to fetch per API request (default: 200, max: 200)
      """
      todoist_client = ctx.request_context.lifespan_context.todoist_client
  
      try:
          logger.info(f"Getting tasks with project_id: {project_id}, section_id: {section_id}, parent_id: {parent_id}, label: {label}, nmax: {nmax}, limit: {limit}")
  
          # Early exit for zero requests to avoid unnecessary API calls
          if nmax is not None:
              if nmax == 0:
                  logger.info("nmax=0 specified, returning empty result")
                  return []
              elif nmax < 0:
                  logger.warning(f"Invalid nmax {nmax}, using default of 100")
                  nmax = 100
  
          if limit > 200:
              logger.warning(f"Limit {limit} exceeds API maximum of 200, using 200 instead")
              limit = 200
          elif limit <= 0:
              logger.warning(f"Invalid limit {limit}, using default of 200")
              limit = 200
  
          # Key optimization: match page size to actual need to reduce API payload
          effective_limit = limit
          if nmax is not None and nmax < limit:
              effective_limit = nmax
              logger.info(f"Optimized limit from {limit} to {effective_limit} to match nmax")
  
          params = {}
          if project_id:
              params["project_id"] = project_id
          if section_id:
              params["section_id"] = section_id
          if parent_id:
              params["parent_id"] = parent_id
          if label:
              params["label"] = label
          if ids:
              params["ids"] = ids
          params["limit"] = effective_limit
  
          tasks_iterator = todoist_client.get_tasks(**params)
          all_tasks = []
          pages_fetched = 0
  
          for task_batch in tasks_iterator:
              pages_fetched += 1
              all_tasks.extend(task_batch)
  
              logger.info(f"Fetched page {pages_fetched} with {len(task_batch)} tasks (total: {len(all_tasks)})")
  
              if nmax is not None and len(all_tasks) >= nmax:
                  # Trim excess - rare due to effective_limit optimization, but handles edge cases
                  all_tasks = all_tasks[:nmax]
                  logger.info(f"Reached nmax of {nmax} tasks, stopping pagination")
                  break
  
              # Todoist API signals end of results by returning fewer items than requested
              if len(task_batch) < effective_limit:
                  logger.info(f"Received {len(task_batch)} tasks (less than limit {effective_limit}), reached end of results")
                  break
  
          if not all_tasks:
              logger.info("No tasks found matching the criteria")
              return "No tasks found matching the criteria"
  
          logger.info(f"Retrieved {len(all_tasks)} tasks total across {pages_fetched} pages")
  
          if nmax is None:
              logger.info("Fetched ALL matching tasks (nmax=None specified)")
          elif len(all_tasks) == nmax:
              logger.info(f"Retrieved exactly the requested {nmax} tasks")
  
          return json.dumps([task.to_dict() for task in all_tasks], indent=2, default=str)
  
      except Exception as error:
          logger.error(f"Error getting tasks: {error}")
          return f"Error getting tasks: {str(error)}"

• src/main.py:85-85 (registration)

  Registers the todoist_get_tasks handler as an MCP tool using FastMCP's decorator.

  mcp.tool()(todoist_get_tasks)

• src/main.py:26-36 (registration)

  Imports the todoist_get_tasks function from tasks module for registration.

  from .tasks import (
      todoist_get_task,
      todoist_get_tasks,
      todoist_filter_tasks,
      todoist_add_task,
      todoist_update_task,
      todoist_complete_task,
      todoist_uncomplete_task,
      todoist_move_task,
      todoist_delete_task,
  )