Command Guide

Generates an AI-powered summary of recent messages in a channel. If you provide all parameters upfront, the summary runs immediately. Otherwise, an interactive wizard walks you through each option.

After generating, you can export the summary as Markdown or PDF. If you have Trello or Notion connected, extracted tasks can be synced directly to your boards.

Search channel messages and get a direct answer with evidence and jump links back to the original messages. Creates a persistent thread so you can ask follow-up questions using the same context.

Upload a .txt, .md, or .pdf file and get an AI summary. You can also reply to a message that has a file attachment. Max file size is 2 MB.

Upload a file and ask a specific question about its content. Creates a thread for follow-up questions with cached context, so subsequent questions are fast.

Display the kanban board for the current channel with three columns: To Do, In Progress, and Done. Each task shows a short hex ID you can use with other commands. Includes interactive buttons to add, move, and remove tasks.

Quickly create a new task. After creation, you can optionally assign it to a server member using the dropdown.

Move a task to the Done column. Use the task's 6-character hex ID shown on the board. Autocomplete helps you find the right task.

When enabled, messages matching task patterns are automatically added to the board. Detected patterns include:

• To Do:, TODO:, Task:, Action Item:
• Reminder:, Note to self:, Need to:
• Unchecked checkboxes: [ ] or - [ ]
• Reply with just todo to turn any message into a task

An interactive wizard walks you through setting up an automated summary:

1. Select your timezone
2. Pick source channel(s) - one, several, or all
3. Choose frequency: daily or weekly
4. Set the time of day (15-minute increments)
5. For weekly: pick the day of the week
6. Choose summary type: overview or tasks
7. Select delivery: channel post or private thread
8. Optionally add a custom focus prompt

Schedules are DST-aware and automatically adjust for daylight saving time changes.

Shows all scheduled summaries for the server including their ID, source channel, destination, frequency, time, and status.

Delete a scheduled summary by its 6-character ID. Shows a confirmation before deleting.

Runs the scheduled summary right now without changing its next scheduled time. Useful for testing or getting an on-demand summary.

Opens a secure OAuth link to authorize SummaryBot with Trello. After connecting, use /integrate trello-board to select which board and list to sync tasks to. Supports two-way sync - tasks flow between Discord and Trello.

Specify which Trello board and list to sync extracted tasks to. Automatically sets up a webhook for inbound sync from Trello.

Opens a secure OAuth link to authorize SummaryBot with Notion. After connecting, use /integrate notion-database to select which database to sync tasks to.

Specify which Notion database to sync extracted tasks to. Validates that the database exists and has the required properties.

Shows which integrations are currently connected and their configuration - account info, target board/database, and sync status.

Disconnects Trello, removes stored credentials, and stops task syncing. Previously synced tasks in Trello are not deleted.

Disconnects Notion, removes stored credentials, and stops task syncing. Previously synced tasks in Notion are not deleted.

Set up a keyword alert and get DM'd whenever that word appears in your server. Uses word-boundary matching (case-insensitive) with a 5-minute cooldown between notifications for the same keyword. DMs include the message context and a jump link.

Shows all active keyword alerts you've set up in the current server.

Remove a keyword alert. Autocomplete helps you find the right keyword.

Set up scheduled email digests with an interactive wizard:

1. Enter your email address
2. Choose frequency: daily, every 12 hours, or weekly
3. Pick source channels: all or specific ones (up to 10)
4. Optionally add a custom focus (e.g. “engineering decisions only”)

Digest emails include key discussions, decisions, tasks, important dates, announcements, and open questions. Your email is masked in status displays for privacy.

Disables your email subscription. Your settings are preserved so you can re-subscribe later without reconfiguring.

Shows your current digest settings: email (masked), frequency, source channels, and custom focus.

Prevent a channel from being included in any summaries, queries, or scheduled reports. Takes effect immediately for all future operations.

Restore a previously excluded channel so it's included in summaries again.

Shows all channels currently excluded from summarization in the server.

Configure the language for all AI-generated output including summaries, task extraction, and UI strings. Supports 98 languages including English, Spanish, French, German, Japanese, Korean, Chinese, Arabic, and many more. Autocomplete helps you find the right language.

Shows the language currently configured for the server.

Shows your server's AI token usage: percentage used, remaining tokens, estimated messages, voice call usage, and reset date. Includes a visual progress bar.

Shows a categorized overview of all commands. Pass a specific command name to get detailed help for that command.

Register a URL to receive POST payloads whenever a summary is generated. Payloads are HMAC-signed and include the summary text, channel info, message count, and timestamp.

Shows all webhooks for the server including their label, URL preview, and enabled/disabled status.

Remove a webhook by its label. Takes effect immediately.

Send a test event to all configured webhooks to verify they're working. Uses the same signing and format as production payloads.

Download recent summaries as Markdown or PDF files. You can filter by channel and choose how many summaries to include (1–50).

Summarize the first .txt, .md, or .pdf attachment on a message without typing any command.

Opens a thread where you can ask questions about a file's content. Follow-up questions are automatically answered using cached context.