How to Connect YNAB to Claude (and What It Can't Do)
YNAB budgeters were early to this. YNAB has offered a clean, well-documented API since 2018, and the community wrapped it in MCP servers almost as soon as the protocol existed — calebl/ynab-mcp-server, chrisguidry/you-need-an-mcp, and Maronato/ynab-mcp are all active options. There’s still nothing official: no YNAB entry in Claude’s connector directory and no first-party server from YNAB as of mid-2026.
The result is a surprisingly capable budget copilot that runs on your say-so or on a clock. Capable because the YNAB API is read-write; limited because nothing in your budget can start it.
From personal access token to talking budget
YNAB’s API uses personal access tokens, which keeps setup honest and simple. In YNAB, open Account Settings → Developer Settings and click New Token. Give that token to whichever community server you chose (most take it as a YNAB_API_TOKEN environment variable), then register the server in Claude via Settings → Connectors as a custom connector — a paid-Claude-plan feature.
The token grants full API access to your budgets, so the usual advice applies doubly for financial data: pick a server whose code you can read, and revoke the token if you stop using it.
Budget conversations that actually work
Because the API covers accounts, categories, and transactions in both directions, Claude can do real data entry, not just reporting:
- “How much is left in Dining Out, and at this month’s pace, when does it hit zero?”
- “I spent $62 at Costco — split it $40 Groceries, $22 Household.”
- “Which categories have I overspent three months running?”
That middle prompt is the sleeper feature. Splitting and recategorizing transactions is the fiddliest part of YNAB’s own UI, and doing it by sentence is genuinely faster. The trend questions are where Claude shines beyond the app entirely — YNAB shows you this month; Claude will happily reason across several.
A budget changes while you’re not looking at it
The catch is architectural. A Cowork scheduled task can point Claude at YNAB each morning, so the “nothing runs unless you’re there” version of this complaint is out of date. What’s left is that YNAB itself can’t reach Claude: nothing in the budget triggers a run.
That matters because budgets fail on specific days. Bank imports land overnight — a morning schedule catches those fine. But a category goes red on the 14th, and the alert that would actually change your behavior has to fire on the 14th, not whenever the timer next comes round. Overspending is an event; a schedule can only approximate it, and the approximation is the difference between course-correcting mid-month and reading about it on the 31st.
ChatGPT Scheduled Tasks and Claude Cowork both run unattended on timers. Their ordinary task schedulers do not subscribe directly to arbitrary events in connected business apps. ChatGPT Workspace Agents can be invoked through an API, but an upstream system must still detect the event and call it.
A simpler way to work with YNAB from email or text
YNAB does not publish a “category just crossed budget” webhook, so no honest integration should promise a truly instantaneous overspending event. The reliable options are on-demand questions and scheduled checks.
Carly provides a native YNAB connection without asking you to host an MCP server. Email or text a transaction—“$62 at Costco, split $40 Groceries and $22 Household”—and Carly can log it. Ask what remains in Dining Out before making a purchase, move money between categories, or request a weekly spending summary. A scheduled Carly workflow can also check YNAB at a fixed interval and email a digest, but that is polling on a clock, not YNAB firing an event.
That distinction makes the pitch smaller but more useful: Carly is the low-friction interface for keeping a budget current from the inbox or phone you already use. AI agents start at $35/month, and steps that do not use AI run free. See the YNAB integration and the full integration catalog.
Self-hosted MCP vs. a managed assistant
| Claude + YNAB MCP | Carly | |
|---|---|---|
| Read budgets, categories, transactions | Yes | Yes |
| Enter, split, recategorize transactions | Yes, in chat (API is read-write) | Yes, from email or text |
| Flag overspending | On the next scheduled check | On demand or on the next scheduled check |
| Morning digest of overnight imports | Yes — scheduled task | Yes |
| Month-end summary without being asked | Yes — scheduled task | Yes |
| Setup | Personal access token + community server + paid Claude plan | Plain-English conversation |
| Pricing | Paid Claude plan | AI agents from $35/mo |
Frequently Asked Questions
Does Claude integrate with YNAB?
Yes, through community MCP servers built on YNAB’s public API — calebl/ynab-mcp-server, you-need-an-mcp, and Maronato/ynab-mcp are the best-known. There’s no official YNAB connector in Claude’s directory as of mid-2026; you add a server as a custom connector on a paid Claude plan.
Can Claude add transactions to YNAB?
Yes. The YNAB API is read-write for transactions, and most community servers expose create and update tools, so you can enter, split, and recategorize spending in plain language — when you ask, or on a schedule you set up in advance.
Which YNAB MCP server should I pick?
They’re all thin wrappers over the same API, so choose on language, maintenance, and how easily you can audit the code. Your personal access token grants full budget access, so open source you can read is the real criterion.
Will Claude warn me when a category goes over budget?
Not as a true event. YNAB does not publish a category-overspent webhook. Claude or Carly can catch it during a scheduled check, and Carly can send the resulting digest by email, but the timing depends on the schedule you choose.
What does a hands-off YNAB routine cost?
With Carly, AI agents start at $35/month, and workflow steps that don’t use AI run free. The exact AI usage depends on how often you ask Carly to categorize, analyze, or summarize the budget.
More: Claude connectors · Can Claude send emails · Claude vs Carly · Claude Cowork alternatives · Best AI agents for productivity · Claude + Amazon · Claude + Wix · Claude + Gumroad
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."
