Skip to main content
The /summarize command collects every message in the current thread and produces a formatted Markdown file. The file includes timestamps, reply chains, author names, and any image attachments — making it easy to archive or share the conversation.

Usage

/summarize [small]

Parameters

small
boolean
default:"true"
When true (the default), message headers are wrapped in <small> tags to create a more compact display. Set to false to use standard-size headers.

Behavior

1

Start the command

Run /summarize inside a thread. The bot immediately replies with an ephemeral message:
Creating summary…
2

Build the summary

The bot collects all thread messages in chronological order and generates a Markdown document. Each message includes:
  • An index number.
  • The author’s display name (server nickname if available).
  • A reply reference if the message is replying to another message (e.g., replying to #3).
  • A timestamp in UTC.
  • The message content.
  • Any image attachments embedded as Markdown image links.
3

Receive the file

The ephemeral message updates to:
Here is your summary!
A .md file is attached with the filename {thread-name} Summary.md.

Output format

The generated Markdown file follows this structure:
# Thread name

Thread started in: #parent-channel-name<br />
Thread started at: Mon, 01 Jan 2024 00:00:00 GMT

<small>1: **username** &mdash; Mon, 01 Jan 2024 00:01:00 GMT</small><br />
Message content here

<small>2: **other-user** (replying to #1) &mdash; Mon, 01 Jan 2024 00:02:00 GMT</small><br />
Reply content here
When small is false, the <small> tags are omitted and headers render at normal size.

Permissions

No special permissions are required. Any server member can run /summarize inside a thread.

Error cases

Running /summarize outside of a thread returns an ephemeral error:
This command can only be used in a thread
If the summary cannot be created (for example, due to a permissions or API error), the ephemeral message updates with the full error details instead of the file.

Build docs developers (and LLMs) love