How to Connect Canny to Claude (and What It Can't Do)
Canny is one of the few feedback tools that treats Claude as a first-class citizen. The company operates an official MCP server at https://api.canny.io/api/mcp/v1 and publishes the exact connect-to-Claude recipe — OAuth client credentials included — right in its help center. No community code, no self-hosting. The server has grown fast since launch, too: per Canny’s changelog, it’s up to 55+ tools spanning ideas, boards, users, companies, and the portal, with recent additions for listing insights and idea comments.
Standard Claude constraints still frame it, as of mid-2026. It installs as a custom connector, so a paid Claude plan is required (Canny’s side needs the Ideas beta, Pro plan and above). And the server answers only when spoken to — inside a chat you started. Votes accumulate and feedback lands around the clock; Claude notices none of it until you ask.
Hooking it up
Following Canny’s help article, the Claude setup is:
- Open claude.ai/settings/connectors and choose Add custom connector.
- Name it “Canny”, set the MCP server URL to
https://api.canny.io/api/mcp/v1, pick OAuth, and paste in the client ID and secret Canny publishes in the help doc. - Click Connect and sign in to your Canny workspace.
A Canny Owner has to make the first connection; teammates can connect afterward, with tool access trimmed to their role. Read tools (list_, retrieve_) are open to everyone connected. Write tools — creating ideas, updating posts, merging duplicates, changing statuses — follow role permissions.
The questions it turns into seconds
The payoff comes when votes, revenue, and comment threads need cross-referencing:
“Which posts on the feature-requests board gained the most votes in the last 30 days?”
“List every open idea from companies above $10k MRR that no admin has responded to.”
“Read the comments on the dark-mode post, summarize the recurring asks, and draft a status update announcing it’s in progress.”
That third one crosses into write territory — with the right role, Claude can post the status change itself. Merging obvious duplicates by conversation is similarly handy.
What the connector sleeps through
A feedback board’s most important moments happen precisely between your chats:
- An idea quietly crosses 100 votes on Thursday — the same-day signal a PM wanted, surfaced whenever someone next thinks to ask.
- A churn-risk enterprise account leaves pointed feedback; nobody’s watching the board that afternoon.
- A feature ships, and the 200 voters on the post hear nothing until a human remembers the status update.
Canny’s MCP has the tools for all three responses. What it lacks — what every Claude connector lacks — is a way to initiate. No vote-threshold triggers, no new-post events, no schedules.
Closing the loop without being in the room
Carly, an AI executive assistant built around triggers, picks up that half:
- Vote-threshold alerts — an idea crosses the line you set, and the PM hears about it in Slack within minutes
- Feedback-to-backlog routing — new high-signal posts get summarized and pushed toward Linear or your tracker
- Ship-day loop-closing — when a feature goes live, the status update and the notification email go out (sent from Gmail or Outlook, not left as drafts) without anyone remembering to do it
Workflows are built conversationally — you say what should happen, Carly asks what “high-signal” means to you, then runs it hosted, 24/7. AI agents start at $35/month, with non-AI workflow steps free and unlimited. More on the Canny integration page and the full integrations catalog.
Chat connector vs. triggered agent
| Claude + Canny MCP | Carly | |
|---|---|---|
| Search and summarize ideas, votes, comments | Yes, in chat | Yes |
| Post status updates, merge duplicates | Yes, when asked (role permitting) | Yes, in a workflow |
| React when a post crosses a vote threshold | No | Yes |
| Watch for feedback from key accounts | No | Yes |
| Notify voters the day a feature ships | Only if you remember to ask | Yes, on the trigger |
| Runs unattended | No | Yes, cloud-hosted |
| Entry cost | Paid Claude plan + Canny Pro (Ideas) | AI agents from $35/mo |
Frequently Asked Questions
Does Claude integrate with Canny?
Yes — officially. Canny operates its own MCP server (api.canny.io/api/mcp/v1) and documents the Claude connection in its help center. You add it as a custom connector, which needs a paid Claude plan; on Canny’s side it’s available to Ideas teams on the Pro plan and above, as of mid-2026.
Can Claude change post statuses or merge feedback in Canny?
Yes, within a chat — the official server includes write tools for creating ideas, updating posts, merging, and status changes, gated by your Canny role. It will never do any of that spontaneously; each action follows an instruction you typed.
Why isn’t Canny in Claude’s connector directory?
As of mid-2026 Canny distributes its server as a manual custom-connector setup rather than a directory listing, so you enter the URL and OAuth credentials yourself. The five-minute walkthrough lives in Canny’s help article.
Can I get an alert when an idea passes a certain vote count?
Not from the Claude integration — connectors can’t watch for events. That’s a trigger workflow, which is what Carly runs: it monitors the board and pings you the moment the threshold is crossed. AI agents start at $35/month.
Which Canny plans include the MCP server?
Canny lists it as available to teams in the Ideas beta on Pro and above. Pair that with a paid Claude plan for the custom connector and you’re set — there’s no separate charge for the server itself.
More: Claude connectors · Can Claude send emails · Claude vs Carly · Claude Cowork alternatives · Claude for product management · Claude + Productboard · Claude + Gorgias · Claude + Freshservice
Ready to automate your busywork?
Carly schedules, researches, and briefs you—so you can focus on what matters.
See what people say
"Before Carly, I relied on a Calendly link, but the whole process felt impersonal and not very professional. Carly changed that by handling all the back-and-forth, so I'm no longer stuck in endless email threads trying to line up schedules.
Now Carly reaches out to candidates, shares my real-time availability, lets them pick a slot, then sends a Zoom link and drops it straight into my calendar. She sends reminders to both of us before each call, which has significantly reduced no-shows and last-minute confusion.
On top of scheduling, Carly acts like a full executive assistant, sending me my schedule the night before so I can prepare for each call. It reminds me of the old x.ai assistant, but Carly is noticeably smarter, faster, and better suited to my healthcare recruitment business."
