Provider REST API

Communication

Connect Claude to Discord

Read and post in a Discord server's channels, and look up its members, roles and threads. Toolspoke puts 12 of its actions behind one MCP endpoint that Claude, Cursor, and Codex all speak.

Connection
Provider REST API
Authentication
Bot token
Actions exposed
12
Cost per call (typical)
1 credit
Adapter
Maintained by Toolspoke

Connected in three steps

  1. 1

    Install Discord

    Open the marketplace in your workspace, add Discord to the project your agents work in, and it appears on the gateway immediately.

  2. 2

    Connect the credential

    Authenticate with bot token. Where to get one, and what it has to be able to reach, is the next section.

  3. 3

    Point your agent at the gateway

    Give your client one address, https://toolspoke.com/mcp. Claude Code takes it as a command, Claude and Claude Desktop add it as a custom connector, and Cursor, Codex and VS Code each read it from a config file of their own.

.mcp.json
{
  "mcpServers": {
    "toolspoke": {
      "type": "http",
      "url": "https://toolspoke.com/mcp"
    }
  }
}

One block covers every tool you have installed. Discord shows up in the client as soon as your policy allows it, and so does everything else you install later.

Where the address goes, per client

Claude Code

Run it in your project, then /mcp to sign in

claude mcp add --transport http toolspoke https://toolspoke.com/mcp
Claude and Claude Desktop

Settings, then Connectors, then Add custom connector

https://toolspoke.com/mcp
Cursor

~/.cursor/mcp.json, or .cursor/mcp.json for one project

{ "mcpServers": { "toolspoke": { "url": "https://toolspoke.com/mcp" } } }
Codex

~/.codex/config.toml

[mcp_servers.toolspoke]
url = "https://toolspoke.com/mcp"
VS Code

.vscode/mcp.json, or the MCP: Add Server command

{ "servers": { "toolspoke": { "type": "http", "url": "https://toolspoke.com/mcp" } } }

What Discord asks for

Bot token. You provide it once, when you install the connector. Toolspoke encrypts it at rest and decrypts it only for the length of a single call, and the gateway attaches it to the outbound request itself, so it is never part of the arguments an agent sends.

Bot tokenRequired
Discord Developer Portal → your application → Bot → Reset Token. Sent as `Authorization: Bot <token>`; a token pasted without its Bot prefix is fine, the prefix is added here.
MTIzNDU2Nzg5MDEyMzQ1Njc4.Abcdef.…
Server (guild) IDRequired
Right-click the server in Discord with Developer Mode on → Copy Server ID. It is the server every guild-level action defaults to, and the one the connection is tested against.
1200000000000000000

What Claude can do in Discord

12 actions, each one declared and named by the connector rather than discovered at runtime. A workspace policy grants a person all of them, a hand-picked selection, everything on the read side, everything on the write side, or none.

Reads
7Reads
Writes
2Writes
Destructive
3Destructive

Reads

7

Fetches data and changes nothing.

  • list_guild_channels

    List every channel in the server the bot can see, with each one's id, name, type, topic, category and position. This is the first call to make: channel ids are what every message operation takes, and Discord's own UI never shows them without Developer Mode. Threads are not included here - Discord returns them from a separate active-threads endpoint this connector does not expose.

  • get_channel

    Fetch one channel or thread by id: its name, type, topic, slow-mode setting, parent category and, for a thread, its archive state and message count. Use it to confirm what a channel id actually points at before posting into it.

  • list_messages

    Read recent messages from a channel or thread, newest first. Pass `before` with a message id to page backwards through history, `after` to read forward from a point, or `around` to read either side of one message; the three are mutually exclusive and Discord ignores all but one. The bot needs both View Channel and Read Message History on the channel - without the second, this returns an empty list rather than an error.

  • get_message

    Fetch one message in full by id: its content, author, timestamps, attachments, embeds, reactions and, if it is a reply, the message it refers to. Use it after list_messages when you need everything about a single message.

  • list_guild_members

    List the server's members, with each one's user record, nickname, roles and join date, ordered by user id. IMPORTANT: this endpoint requires the SERVER MEMBERS privileged intent to be enabled for the application in the Discord developer portal. Without it Discord answers 403, and the message reads like the bot's role is missing a permission rather than like a switch in a different console. Page with `after`, passing the highest user id from the previous response.

  • get_user

    Fetch one user's public account record by id: username, global display name, discriminator, avatar and whether the account is a bot. This is the global user, not the server member - nickname and roles come from list_guild_members. Pass the bot's own id to confirm which application this token belongs to.

  • list_guild_roles

    List the server's roles with their ids, names, colours, permission bitfields and position. Use it to find the role ids that send_message takes as mention_role_ids, and to read what a member's role list from list_guild_members actually means.

Writes

2

Creates or updates something on the other side.

  • create_thread

    Start a thread in a text channel. Classified a write rather than destructive: it creates an empty container that carries no message of its own and can be deleted afterwards by anyone with Manage Threads. IMPORTANT: Discord's own default for this endpoint is a PRIVATE thread, kept for backwards compatibility, so a thread created without naming a type is invisible to the channel. This connector defaults `type` to 11, a public thread, which is what almost every caller means. Post into the thread by calling send_message with the id this returns.

  • add_reaction

    React to a message as the bot. A write rather than destructive: a reaction carries no text, changes nothing anyone has read, and is removed as easily as it is added. `emoji` takes a plain Unicode character for a standard emoji, or `name:id` for a custom server emoji - the id half is what makes it resolve, and a custom emoji sent by name alone is rejected. Discord answers 204 with no body.

Destructive

3

Deletes or permanently alters something. Worth granting on purpose.

  • send_message

    Post a message into a channel or thread. Classified destructive because every human watching that channel sees it immediately and there is no taking it back - a write grant does not reach this. Mentions are inert by default: an @everyone or @user written in `content` renders as text and notifies nobody unless mention_parse, mention_user_ids or mention_role_ids says otherwise. Set reply_to_message_id to post it as a reply to an existing message.

  • edit_message

    Rewrite a message this bot posted. Discord does not allow a bot to edit anyone else's message, so this only ever works on its own. Classified destructive for the same reason send_message is: the new text replaces what people have already read, the old text is gone, and an edit is visible in the channel. Fields left out are unchanged; pass content to replace the text.

  • delete_message

    Delete a message. There is no undo and no API that returns a deleted message, so this is destructive in the plainest sense. Deleting the bot's own message needs no extra permission; deleting anyone else's needs Manage Messages on the channel. Discord answers 204 with no body.

What it will not do

Enforced by the gateway rather than left to convention, which is why each of these can be stated flatly.

It cannot call anything else
The 12 actions above are the whole of it. A call to any other name is refused before it reaches Discord rather than forwarded on, and connecting your account does not add to the list: it is fixed by the connector, not discovered at run time.
It reaches no further than your credential
Toolspoke holds no access to Discord of its own. Every call carries the credential you stored and nothing besides, so whatever that credential cannot reach, this connector cannot reach either.
It never hears from Discord
Nothing is pushed to it. There is no webhook, no subscription and no polling, so this connector cannot notice by itself that something changed in Discord. An agent has to ask.
It does not smooth over provider limits
Toolspoke does not retry, queue or back off around Discord's own rate limits. A call that Discord refuses comes back to the agent as a failed call.

Before you connect it

What can Claude do in Discord?

12 named actions: 7 that only read, 2 that write and 3 that delete or permanently alter something. They include list_guild_channels, get_channel and list_messages. Nothing outside that list is reachable: the connector declares each operation by name rather than proxying whatever an agent asks for.

What credentials does the Discord connector need?

Bot token. The connector asks for bot token and server (guild) id. Values are encrypted at rest and attached to the outbound request by the gateway, so they are never part of the arguments an agent sends and never reach the audit log.

Does the Discord connector work with Cursor and Codex, or only Claude?

Any client that speaks MCP, and every one of them gets the same 12 actions. There is a single address, https://toolspoke.com/mcp. Claude Code adds it with claude mcp add --transport http, Claude and Claude Desktop take it as a custom connector in settings, Cursor reads it from .cursor/mcp.json, Codex from ~/.codex/config.toml, and VS Code from .vscode/mcp.json. Each of them signs in to the gateway itself, so there is no key to paste.

What does the Discord connector not do?

The 12 actions above are the whole of it. A call to any other name is refused before it reaches Discord rather than forwarded on, and connecting your account does not add to the list: it is fixed by the connector, not discovered at run time. Toolspoke holds no access to Discord of its own. Every call carries the credential you stored and nothing besides, so whatever that credential cannot reach, this connector cannot reach either. Nothing is pushed to it. There is no webhook, no subscription and no polling, so this connector cannot notice by itself that something changed in Discord. An agent has to ask. Toolspoke does not retry, queue or back off around Discord's own rate limits. A call that Discord refuses comes back to the agent as a failed call.

Can I limit which actions an agent can call?

Yes, in two places. The project switches Discord's actions on and off one at a time, for everyone in the project at once, and the screen groups them by read, write and destructive so turning off everything that deletes is one click. An individual agent key can then be narrowed further, to particular toolkits in a project and to particular actions in a toolkit. Whatever it was granted, a key never reaches a project its owner cannot.

What gets recorded when an agent calls Discord?

Every attempt, with the agent that made it and the person that agent belongs to, the full request payload, the response payload, the status, the duration, and the credits spent. Values whose key names a secret are masked out before the record is shown to anyone. An operation the connector marks as not retained never has its response body written at all, so the gateway keeps no second copy of what was read.