token configuration and a room_id input. Set up the token before running a flow.
Choose a Chatwork operation
CHATWORK_GET_MESSAGES accepts optional Boolean force: true fetches up to 100 latest messages; the default false fetches only new messages. CHATWORK_GET_TASKS can filter by account_id, assigned_by_account_id, or status (open or done). CHATWORK_CREATE_TASK accepts optional limit (a Unix timestamp or YYYY/MM/DD date) and limit_type (none, date, or time).
Send a Chatwork message
result.message_id, the ID of the sent message. If the step fails, check that the token is valid, the bot or user can access the room, and the room_id is correct. The input field is message, not body; the configuration field is token, not api_token.
Send a dynamic report to Chatwork
A message can combine results from earlier steps. This keeps the weekly report example from the former advanced guide; replace the step IDs and result fields with outputs from your flow.Create and complete a Chatwork task
to_ids to a comma-separated string of account IDs, such as "6454321,7654321".
To complete the task in a later step, use the ID returned by create_task:
result.task_id to verify which task changed.
Review open tasks
CHATWORK_GET_TASKS can retrieve open tasks for a room, then pass them to a later step for a status report. Use account_id to limit the report to one assignee when needed.
limit_time (a Unix timestamp when a due time is set). Use the returned list to prepare a reminder or report; an empty array means there are no matching tasks.
To monitor deadlines, compare each set limit_time with the current time in a Python step and group open tasks as overdue or upcoming. limit_time is a Unix timestamp, so convert it to a date before displaying it to people.
Review recent messages
UseCHATWORK_GET_MESSAGES to fetch new messages since the last fetch, or set force to true to fetch up to 100 latest messages. The result is a message array with each message’s body, sender account, and send and update timestamps. You can pass this array to a later analysis step to identify topics, questions, and action items.
steps.get_recent_messages.result. You can analyze it and post a dynamic report to the same room. This example summarizes the latest messages with OPENAI_INVOKE; configure OpenAI access or Jinba credit before running it.