Slash Commands and Interactions #2295
Conversation
Co-authored-by: Advaith <advaithj1@gmail.com>
msciotti
changed the title
|
This is my final draft now. All the information should be correct, even if some could be better. I'll be merging and deploying this tonight so after that any changes can be PR'd against master. |
|
|
||
| Fetch all of the global commands for your application. Returns an array of [ApplicationCommand](#DOCS_INTERACTIONS_SLASH_COMMANDS/applicationcommand) objects. | ||
|
|
||
| ## Create Global Application Command % POST /applications/{application.id#DOCS_TOPICS_OAUTH2/application-object}/commands |
There was a problem hiding this comment.
Is there a limit to the payload size for this endpoint? (Hopefully it's large enough, the previous permission example was 1kb minified 😰)
There was a problem hiding this comment.
A limit to the payload size for creating commands?
There was a problem hiding this comment.
Like how there's limits to embeds and whatnot, I guess. I was mostly just asking for the sake of asking, though.
There was a problem hiding this comment.
Those limits are mostly just because our send_message has a max filesize on it I believe
| | tts? | bool | is the response TTS | | ||
| | content | string | message content | | ||
| | embeds? | array of embeds | supports up to 10 embeds | | ||
| | allowed_mentions? | allowed mentions | [allowed mentions](#DOCS_RESOURCES_CHANNEL/allowed-mentions-object) object | |
There was a problem hiding this comment.
From what i've seen, flags would be here too.
There was a problem hiding this comment.
I've intentionally left flags off for now because ephemeral messages aren't quite done (UI-wise) on all platforms.
|
|
||
| ## Edit Followup Message % PATCH /webhooks/application.id/{interaction.token#DOCS_INTERACTIONS_SLASH_COMMANDS/interaction}/messages/{message.id#DOCS_RESOURCES_CHANNEL/message-object} | ||
|
|
||
| Edits a followup message for an Interaction. Functions the same as [Edit Webhook Message](#DOCS_RESOURCES_WEBHOOK/edit-webhook-message). |
There was a problem hiding this comment.
Does this do the same behavior as setting wait to true in executing commands?
(meant to put this in creating a follow-up message)
There was a problem hiding this comment.
Not entirely sure what you mean. Do you mean will it return the message object when creating a followup message?
There was a problem hiding this comment.
Yeah, that was my question. Meant to say executing webhooks not commands.
| | data?\* | ApplicationCommandInteractionData | the command data payload | | ||
| | guild_id | snowflake | the guild it was sent from | | ||
| | channel_id | snowflake | the channel it was sent from | | ||
| | member | GuildMember | guild member data for the invoking user | |
There was a problem hiding this comment.
Is this guaranteed to have the User attached, unlike CREATE_MESSAGE?
There was a problem hiding this comment.
Yes the user object is guaranteed
|
|
||
| After receiving an interaction, you must respond to acknowledge it. This may be a `pong` for a `ping`, a message, or simply an acknowledgement that you have received it and will handle the command async. | ||
|
|
||
| Interaction responses may choose to "eat" the user's command input if you do not wish to have their slash command show up as message in chat. This may be helpful for slash commands, or commands whose responses are asynchronous or ephemeral messages. |
There was a problem hiding this comment.
I feel "swallow" is the more widely used term here, but "eat" is a fun word.
There was a problem hiding this comment.
Yknow, I knew "eat" wasn't quite it but I knew it was some sort of eating-based word. I'm keeping it now though I've become attached :p
| @@ -39,6 +39,13 @@ Many of these libraries are represented in the [unofficial, community-driven Dis | |||
| | [Sword](https://github.com/Azoy/Sword) | Swift | | |||
| | [Discordeno](https://github.com/Skillz4Killz/Discordeno) | TypeScript | | |||
|
|
|||
| ## Interactions | |||
|
|
|||
| [Interactions and Slash Commands](#DOCS_INTERACTIONS_SLASH_COMMANDS/) are the great, new way of making a Discord bot. The following open-source libraries provide help for the security and authentication checks that are mandatory if you are receiving Interactions via outgoing webhook. They also include some types for the Interactions data models. | |||
There was a problem hiding this comment.
The comma after great feels weird.
|
|
||
| ## Create Followup Message % POST /webhooks/application.id/{interaction.token#DOCS_INTERACTIONS_SLASH_COMMANDS/interaction} | ||
|
|
||
| Create a followup message for an Interaction. Functions the same as [Execute Webhook](#DOCS_RESOURCES_WEBHOOK/execute-webhook) |
There was a problem hiding this comment.
Can this use different interaction types (like ChannelMessageWithSource) or is it completely restricted to the parameters from webhooks? If it is restricted, will we be able to use file in interaction responses or would it only be here?
There was a problem hiding this comment.
Followups are only webhook messages, types are only for the initial response
There was a problem hiding this comment.
Advaith is correct that only the original response supports the different types since any followup message is just like sending a normal message.
I don't believe we support file yet, but we should. Had to make some tradeoffs in the architecture to start but definitely noted :)
| | INTEGER | 4 | | ||
| | BOOLEAN | 5 | | ||
| | USER | 6 | | ||
| | CHANNEL | 7 | |
There was a problem hiding this comment.
Is type 7 here any channel, only text channels or voice channels as well?
There was a problem hiding this comment.
Any channel - from text channels to channel categories.
There was a problem hiding this comment.
Any channel within the guild, yes. So no DM channels but everything else including categories, store channels, etc.
| @@ -386,6 +387,7 @@ Returns the bot's OAuth2 application info. | |||
| | primary_sku_id? | snowflake | if this application is a game sold on Discord, this field will be the id of the "Game SKU" that is created, if exists | | |||
| | slug? | string | if this application is a game sold on Discord, this field will be the URL slug that links to the store page | | |||
| | cover_image? | string | if this application is a game sold on Discord, this field will be the hash of the image on store embeds | | |||
| | flags | int | the application's public flags | | |||
There was a problem hiding this comment.
Is there any specific reason why this isn't called public_flags?
There was a problem hiding this comment.
I think that all the flags on an application actually are public, as opposed to users where we specifically differentiate. Calling them the public flags covers my butt for now but I can confirm.
There was a problem hiding this comment.
It would be useful for future-proofing for a future where bots could be boosted.
|
Thanks everyone for your help in getting this PR into a good state! As always if we find more things that need clarifying or changing, PRs or pings are always welcome. There's also two new issue templates here for slash command bugs and feature requests, so we'll keep an eye on those! |
|
Mason, could you tell us what the rollout plans are for this feature? Now this is merged and we have feature/bug templates, are we going to see the rollout starting today? |
|
broken link? |
Opening the PR now because the "meat" of the docs are done. Just a quick note:
Comments on this PR should be solely about the documentation. When this launches, we will have specific repo templates for feedback on the feature itself. I will pretty aggressively prune the convo here to be documentation-only so that we can get this out
This documentation page is a bit of a monolith, and should probably be split up into smaller sections. The same can be said of most of our documentation, to be honest. I'm prioritizing having the documentation ready over having it look pretty, and we can revisit later.
TODO:
INTERACTION_CREATEgateway event