Connectors · Developer & infrastructure
Discourse MCP server for AI agents.
actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Discourse and the rest of your systems. It gives your agent 91 Discourse actions (37 reads, 42 writes and 12 that wait for a confirmation) and holds the Discourse credential itself, so the agent never does. It paces Discourse's limits and records every call.
100,000 free calls a month. No card, no email, no password: press Start a free sandbox on the next screen and keep the key it shows you.
claude mcp add --transport http actionlake https://connect.actionlake.com/mcpdo { action: "discourse.admin_users.list", intent: "check which Discourse account is connected"}# List users · read · 1 upstream call- 91actionsin one signed file
- 37readlook, never change
- 42writecreate and update
- 12wait for a yesdestructive, confirmed first
Set up
Connect Discourse in three steps.
- 01
Add actionlake to Claude.
Run the command once in Claude Code; its consent page can create a free sandbox: 100,000 calls a month, no card. In the Claude app, add a custom connector with the URL below and paste your key on the consent page. Creating a new sandbox from the Claude app is not yet tested by us.
https://connect.actionlake.com/mcp - 02
Connect Discourse.
Header credentials. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead. actionlake stores what you paste write-only, sealed under your tenant's key.
api_keyDiscourse API key (Admin > API > New API Key)api_usernameApi-Username: the Discourse username the key acts as (for admin endpoints, an admin such as `system`)hostWhere your Discourse lives: your host
account { op: "connections.requirements", service: "discourse" } # free: exactly what to paste, and what is found for you - 03
Make the first call. Check before a write.
Ask in plain words; Claude picks the action. Before anything changes,
checkvalidates the call and prices it, for free.do { action: "discourse.admin_users.list" }# list users · read · 1 upstream callcheck { action: "discourse.users.create", input: { … } }# free: inputs, permissions, cost and confirmations, nothing sent
Discourse at a glance
- Sign-in
- Header credentials. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead.
- You provide
api_keyapi_username- Setup fields
host- Limits we pace to
- 60 calls a minute, per connection
- API covered
- Discourse API (OpenAPI 3.1, docs.discourse.org/openapi.json, version `latest`), at your Discourse instance's API base URL, which you enter when connecting
- Built from
- Discourse's documentation, read 23 October 2026docs.discourse.orgmeta.discourse.org
- Service file
- Version 1.0.0, signed (Ed25519)
- Cached reads
- 37 of 37 reads, for up to 5 minutes; a write clears what it affects
- Shared entities
- 4, so ask can line Discourse up with your other systems
- Data
- Processed and stored in Sydney, Australia. No US region yet
Capabilities
What an AI agent can do in Discourse.
91 actions in 23 areas. Every one carries its effect, so a key can allow the reads and hold the rest.
- 37readlook, never change
- 42writecreate and update
- 12destructivewait for a confirmation
Topics
124 read · 7 write · 1 destructive
Get the latest topics · Get a single topic · Update a topic · Remove a topic
Users
124 read · 7 write · 1 destructive
Creates a user · Get a single user by username · Update a user · Send password reset email
Admin users
113 read · 3 write · 5 destructive
List users · Delete a user · Merge a user · List users by flag
Posts
73 read · 3 write · 1 destructive
List latest posts across topics · Creates a new topic, a new post, or a private message · Update a single post · delete a single post
Groups
63 read · 2 write · 1 destructive
List groups · Create a group · Get a group · Update a group
Uploads
60 read · 5 write · 1 destructive
Generates batches of presigned URLs for multipart parts · Completes a direct external upload · Complete multipart upload · Creates a multipart external upload
Categories
53 read · 2 write · 0 destructive
Creates a category · Retrieves a list of categories · List topics · Show category
Badges
41 read · 2 write · 1 destructive
Create badge · List badges · Update badge · Delete badge
15 more areas · 28 actions
- Tag groups 4
- Backups 3
- Group members 3
- Invites 2
- Notifications 2
- Private messages 2
- Site 2
- Tags 2
- User badges 2
- Directory items 1
- Events 1
- Post actions 1
- Search 1
- Topic timers 1
- User actions 1
Top actions
The Discourse API, shaped for agents.
The ids your agent types, what each does, and the one name it shares with other systems. 24 of Discourse's 25 action names are shared with at least one other connector, so an agent learns them once.
| Action | What it does | Effect | Same name in |
|---|---|---|---|
discourse.admin_users.listList users | List users. | read | identity.user.list474 connectors |
discourse.groups.listList groups | List groups. | read | identity.group.list262 connectors |
discourse.users.createCreates a user | Creates a user. | write | identity.user.create262 connectors |
discourse.groups.createCreate a group | Create a group. | write | identity.group.create205 connectors |
discourse.admin_users.deleteDelete a user | Delete a user. | destructive | identity.user.delete253 connectors |
discourse.tags.listGet a list of tags | Get a list of tags. | read | crm.tag.list65 connectors |
discourse.group_members.listList group members | List group members. | read | identity.group_member.list103 connectors |
discourse.group_members.addAdd group members | Add group members. | write | identity.group_member.add72 connectors |
discourse.invites.createCreate an invite | Create an invite. | write | identity.invitation.create41 connectors |
discourse.posts.listList latest posts across topics | List latest posts across topics. | read | service.forum_post.list2 connectors |
discourse.posts.createCreates a new topic, a new post, or a private message | Creates a new topic, a new post, or a private message. | write | service.forum_post.create2 connectors |
discourse.topics.list_latestGet the latest topics | Get the latest topics. | read | service.forum_topic.list2 connectors |
Try it
Prompts to paste into Claude.
Each one names the Discourse actions it will call, so you know what it reads and what it changes before you send it.
- Readread
Who has access in Discourse? List every user and their role.
calls
discourse.admin_users.list - Writewrite
Creates a user in Discourse. Show me exactly what you will send before you save it.
calls
discourse.users.create
Governance for Discourse
Destructive actions wait for a yes.
12 Discourse actions are destructive. These are all of them:
- Delete a user
discourse.admin_users.delete - Delete a group
discourse.groups.delete - Remove group members
discourse.group_members.remove - Merge a user
discourse.admin_users.merge - delete a single post
discourse.posts.delete - Remove a topic
discourse.topics.delete - Suspend a user
discourse.admin_users.suspend - Change password
discourse.users.reset_password - Anonymize a user
discourse.admin_users.anonymize - Deactivate a user
discourse.admin_users.deactivate - Delete badge
discourse.badges.delete - Abort multipart upload
discourse.uploads.abort_multipart
A destructive action runs only with confirm: true. Under the human policy the agent cannot confirm for itself: the run stops until a named person approves, in the console or from a signed link in an email. The cap on a key is refused before the call is sent, never after.
Delete a user?
intent: the user asked me to delete a user
account { op: "keys.create", name: "discourse-agent", allowed_actions: ["discourse.*"], cap_calls: 2000, confirm_policy: "human_for_writes", approval_rules: [ { match: "discourse.admin_users.delete", approver: "[email protected]" }, { match: "discourse.groups.delete", approver: "[email protected]" }, { match: "discourse.group_members.remove", approver: "[email protected]" } ]}Cost
What Discourse costs on actionlake.
Every Discourse action is one upstream call, the only thing actionlake bills. Searching, check, cost, account and any call Discourse refused with 401, 403, 429 or a 5xx are free. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.
- 1upstream call per actionall 91 Discourse actions
- 91actions pageeach page of a long list is one call
- 37reads can be cacheda cache hit costs nothing
- 100,000calls a month, freepaid plans from US$49 a month for 250,000
Action index
Every Discourse action.
91 actions by area: the id your agent calls, what it does, and its effect.
Topics12
discourse.topics.bookmarkBookmark topicwritediscourse.topics.deleteRemove a topicdestructivediscourse.topics.getGet a single topicreaddiscourse.topics.invite_groupInvite group to topicwritediscourse.topics.invite_userInvite to topicwritediscourse.topics.list_latestGet the latest topicsreaddiscourse.topics.list_postsGet specific posts from a topicreaddiscourse.topics.list_topGet the top topics filtered by periodreaddiscourse.topics.set_notification_levelSet notification levelwritediscourse.topics.updateUpdate a topicwritediscourse.topics.update_statusUpdate the status of a topicwritediscourse.topics.update_timestampUpdate topic timestampwrite
Users12
discourse.users.createCreates a userwritediscourse.users.getGet a single user by usernamereaddiscourse.users.get_by_external_idGet a user by external_idreaddiscourse.users.get_by_provider_external_idGet a user by identity provider external IDreaddiscourse.users.get_emailsGet email addresses belonging to a userreaddiscourse.users.refresh_gravatarRefresh gravatarwritediscourse.users.reset_passwordChange passworddestructivediscourse.users.send_password_resetSend password reset emailwritediscourse.users.updateUpdate a userwritediscourse.users.update_avatarUpdate avatarwritediscourse.users.update_emailUpdate emailwritediscourse.users.update_usernameUpdate usernamewrite
Admin users11
discourse.admin_users.activateActivate a userwritediscourse.admin_users.anonymizeAnonymize a userdestructivediscourse.admin_users.deactivateDeactivate a userdestructivediscourse.admin_users.deleteDelete a userdestructivediscourse.admin_users.getGet a user by idreaddiscourse.admin_users.listList usersreaddiscourse.admin_users.list_by_flagList users by flagreaddiscourse.admin_users.log_outLog a user outwritediscourse.admin_users.mergeMerge a userdestructivediscourse.admin_users.silenceSilence a userwritediscourse.admin_users.suspendSuspend a userdestructive
Posts7
discourse.posts.createCreates a new topic, a new post, or a private messagewritediscourse.posts.deletedelete a single postdestructivediscourse.posts.getRetrieve a single postreaddiscourse.posts.listList latest posts across topicsreaddiscourse.posts.list_repliesList replies to a postreaddiscourse.posts.lockLock a post from being editedwritediscourse.posts.updateUpdate a single postwrite
Groups6
discourse.groups.createCreate a groupwritediscourse.groups.deleteDelete a groupdestructivediscourse.groups.getGet a groupreaddiscourse.groups.get_by_idGet a group by idreaddiscourse.groups.listList groupsreaddiscourse.groups.updateUpdate a groupwrite
Uploads6
discourse.uploads.abort_multipartAbort multipart uploaddestructivediscourse.uploads.batch_presign_multipart_partsGenerates batches of presigned URLs for multipart partswritediscourse.uploads.complete_externalCompletes a direct external uploadwritediscourse.uploads.complete_multipartComplete multipart uploadwritediscourse.uploads.create_multipartCreates a multipart external uploadwritediscourse.uploads.generate_presigned_putInitiates a direct external uploadwrite
Categories5
discourse.categories.createCreates a categorywritediscourse.categories.getShow categoryreaddiscourse.categories.listRetrieves a list of categoriesreaddiscourse.categories.list_topicsList topicsreaddiscourse.categories.updateUpdates a categorywrite
Badges4
discourse.badges.createCreate badgewritediscourse.badges.deleteDelete badgedestructivediscourse.badges.listList badgesreaddiscourse.badges.updateUpdate badgewrite
Tag groups4
discourse.tag_groups.createCreates a tag groupwritediscourse.tag_groups.getGet a single tag groupreaddiscourse.tag_groups.listGet a list of tag groupsreaddiscourse.tag_groups.updateUpdate tag groupwrite
Backups3
discourse.backups.createCreate backupwritediscourse.backups.listList backupsreaddiscourse.backups.send_download_emailSend download backup emailwrite
Group members3
discourse.group_members.addAdd group memberswritediscourse.group_members.listList group membersreaddiscourse.group_members.removeRemove group membersdestructive
Invites2
discourse.invites.createCreate an invitewritediscourse.invites.create_multipleCreate multiple inviteswrite
Notifications2
discourse.notifications.listGet the notifications that belong to the current userreaddiscourse.notifications.mark_readMark notifications as readwrite
Private messages2
discourse.private_messages.listGet a list of private messages for a userreaddiscourse.private_messages.list_sentGet a list of private messages sent for a userread
Site2
discourse.site.getGet site inforeaddiscourse.site.get_basic_infoGet site basic inforead
Tags2
discourse.tags.listGet a list of tagsreaddiscourse.tags.list_topicsGet a specific tagread
User badges2
discourse.user_badges.createGrant a badge to a userwritediscourse.user_badges.listList badges for a userread
Directory items1
discourse.directory_items.listGet a public list of usersread
Events1
discourse.events.listList calendar eventsread
Post actions1
discourse.post_actions.createLike a post and other actionswrite
Search1
discourse.search.querySearch for a termread
Topic timers1
discourse.topic_timers.createCreate topic timerwrite
User actions1
discourse.user_actions.listGet a list of user actionsread
Questions
The Discourse MCP server, answered.
Is there a Discourse MCP server for Claude?
Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 91 Discourse actions across topics, users, admin users and posts, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Discourse.
What do I need to connect Discourse?
Discourse API key (Admin > API > New API Key) and Api-Username: the Discourse username the key acts as (for admin endpoints, an admin such as `system`). Plus your host. The free call account { op: "connections.requirements", service: "discourse" } lists exactly what to paste.
Can an AI agent delete or change things in Discourse?
Only if its key allows it, and only with an explicit confirmation. 12 Discourse actions are destructive, including delete a user, delete a group and remove group members. Each runs only with confirm: true. Under the human confirm policy, or an approval rule on the key, the run waits until a named person approves it.
Will an AI agent use up my Discourse API limit?
actionlake paces every call to the limits in the Discourse service file: 60 calls a minute per connection. Repeated reads are answered from cache (37 Discourse actions can be), which never reaches Discourse.
What does a Discourse action cost?
One upstream call, the only unit actionlake bills. A list that pages bills one call per page (91 Discourse actions page). Cache hits, searches, dry runs and calls Discourse refused with 401, 403, 429 or a 5xx are free. The Free plan includes 100,000 calls a month with no card. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.
Which Discourse API does it cover?
The Discourse API (OpenAPI 3.1, docs.discourse.org/openapi.json, version `latest`) at your Discourse instance's API base URL, which you enter when connecting. It is built from Discourse's documentation as read on 23 October 2026. 91 actions in 23 areas, each one in the signed service file (version 1.0.0).
Does actionlake keep my Discourse data, and where?
actionlake processes and stores data in Sydney, Australia; there is no US region yet. Reads may be cached for up to 5 minutes, and a write clears the reads it affects. Each run's request and response is kept, sealed, for your plan's trace window: 1 day on Free, 3 days on Starter, 7 days on Pro, 14 days on Team, 30 days on Business and 30 days on Enterprise. Your Discourse credential is sealed under your tenant's key, never logged and never returned. Your traffic is not used to train models.
Which AI clients can use it?
Claude Code, tested by us end to end, and the Claude app as a custom connector, tested by us with an existing key. Other MCP clients that support remote servers over Streamable HTTP (MCP revisions 2025-06-18 and 2025-11-25) and OAuth dynamic client registration, or a bearer header, should connect the same way; we have not tested them. Or plain REST, under the same action names.
Give your agent Discourse today.
Start free - 100,000 calls/month, no card. No signup. Hard caps on every key. Cached answers are free.
