Can an AI agent use OnceHub?
OnceHub is scheduling software for booking meetings.
Rank among the 246 products we drove
- Steps on its own
- tied #3
- 4/ 5
- Find to Leave, no person
- Core work done
- tied #1
- 100%
- 10 of 10 core tasks
- Time to get in
- tied #42
- 4minutes
- Getting in on its own
Yes. Our agent signed up for OnceHub on its own in a browser with an email and password, got an API key and used its API and MCP server. No CLI found.
The record
Level confirmed
Not yet
Level from one drive; a second must agree.
- Got in by
- Sign-up in a browser4 minutes
- Got as far as
- Use: all 10 core tasksNo payment made
Connects by
MCP, API
Agent payments
None found
On its own · 598 doc pages read
Docs it can act on: llms.txt, an OpenAPI and an MCP server.
On its own · 4 minutes
Sign-up in a browser.
On its own · 10 of 10 core tasks
Did: list free time slots over the API and the key-based MCP server; book, cancel and ask guests to reschedule meetings over the API.
Not tried
No payment made.
On its own
Nothing to cancel. It deletes its account.
- On its own
- Not tried
To reach L5: Pay without a person.
See how it ranks among scheduling APIs for agents.
Own OnceHub? 6 things to fix. Get your findings, free
Checked 17 Nov 2026: MCP server asks for a sign-in, sign-up page, llms.txt and OpenAPI unchanged since the check before.
Find
How an agent connects
MCP, API, OpenAPI and llms.txt found.
MCP server
2 of 13 testedOfficial server. Platform MCP mcp.oncehub.com/sse: API key, deprecated HTTP+SSE transport, 2 tools when driven. User MCP my-mcp.oncehub.com: OAuth with DCR, 11 tools in docs. Sign‑in needed to see its tools.
13 tools mapped in our drive
API
30 of 31 testedREST v2, 31 operations, OpenAPI 3.1.0 at help.oncehub.com/openapi.json; API-Key header.
31 operations mapped in our drive
OpenAPI
YesA machine-readable description of the API.
CLI
None foundSDK
None foundllms.txt
YesA file that points agents to the docs.
Agent payments
None foundChose a plan, but the checkout didn't open: Schedule, $12 a month. No CAPTCHA at checkout.
Docs
In partCovers getting started, auth and error codes.
Code examples on only some pages.
Describes each operation only in part: completeness 64 of 100.
7 of 11 documented ways to do the core work were tried, and 7 of them worked.
598 of 852 pages read
Get in
Getting in
1 of 2 ways in worked without a person.
The way that worked
Sign-up in a browser4 minutes
- An inboxNeeded
- A browserNeeded
- A personNot needed
- Agent on a serverAPI, MCP, CLINoStops at sign-up: it works only in a browser.
- Assistant in its own browserLike ChatGPT agentPartlySigns up, confirms its email and does all of the core work. Payment wasn't tried.
- Signed agentKnown to CloudflareYesThe sign-up page opens.
Every way in
| Way in | What happened |
|---|---|
| Sign-up in a browser | Worked |
| Sign-in for apps (OAuth) | Stopped |
- CrawlersThe robots.txt on account.oncehub.com asks crawlers to stay off sign-up. Nothing to fix.
- Bot checkGoogle reCAPTCHA let our agent through at sign-up. Our agent signs every request (Web Bot Auth).
Use
What an agent can do with OnceHub
All 10 core tasks done, and 17 of 29 others.
Core tasks
- done there
- offered, not tried
| Task | Result | Through |
|---|---|---|
| Create a booking calendar | Done | API: done there |
| Book a meeting on a booking calendar | Done | API: done thereMCP: done there |
| List open time slots of a booking calendar | Done | API: done thereMCP: done there |
| Cancel a booking | Done | API: done thereMCP: offered, not tried |
| Reschedule a booking or ask the guest to reschedule | Done | API: done thereMCP: offered, not tried |
| Set scheduling availability | Done | API: done thereMCP: offered, not tried |
| Group booking calendars in a booking hub | Done | MCP: offered, not tried |
| Build a routing form | Done | Web app: done there |
| Create a web chatbot | Done | Web app: done there |
| Set up an AI phone agent for booking | Done | Web app: done there |
Surface audit: 59 documented operations: 35 run and 24 blocked by sandbox or plan limits. Of the 35 run, 33 passed and 2 didn't; 9 behaved differently from the docs. By surface, 30 on the API, 2 on the MCP server and 3 elsewhere, such as sign-in and webhooks.
Holds up
84of 100
- Reliability85
- Handles mistakes82
The mean of the two, 83.5, with a half rounded to the even number.
Docs match
71of 100
- Docs accuracy74
- Docs completeness64
Docs accuracy counts 70%, completeness 30%.
Surface tested
56of 100
33 of the 59 documented operations ran and passed
What it can't do, and why
- Create or revoke an API key without the dashboard Keys can only be made and deleted in the web dashboard.
- Keep booking after about ten bookings An undocumented booking limit refused every new booking.
- Add an AI agent with knowledge sources to a chatbot OnceHub staff must switch the feature on.
- Reassign bookings or run round robin A second user needs a seat; the invited user stayed pending.
17 of 29 other tasks done
- List bookingsDone
- Mark a booking as no-showDone
- Create a one-time booking linkDone
- Approve or decline pending meeting requestsDone
- Offer group sessionsDone
- Add questions to the booking formDone
- Customize email and text notifications to guestsDone
- Create a contactDone
- Update a contactDone
- Delete a contactDone
- Import contacts from a fileDone
- Read booking analyticsDone
- Export activities to a fileDone
- Register a webhookDone
- Save a booking calendar templateDone
- Add a user to the accountDone
- Set a user's roleDone
- Embed a booking calendar on a websitePartly
- Reassign a booking to another hostA second user needs a seat; the invited user stayed pending.Blocked
- Distribute meetings round-robin across a teamA second user needs a seat; the invited user stayed pending.Blocked
- Collect payment for a booking with StripeBlocked
- Review phone call recordings and AI summariesBlocked
- Add knowledge sources to a chatbot AI agentOnceHub staff must switch the feature on.Blocked
- Connect an external calendarNot tried
- Connect a video conferencing appNot tried
- Connect a CRM (Salesforce or HubSpot)Not tried
- Configure single sign-onNot tried
- Send notifications from a corporate mailboxNot tried
- Brand the booking pageNot tried
Needs a paid plan
- Set up an AI phone agent for bookingpaid subscription (phone licenses)
The drive used the no-card trial.
Tips for agent builders
- Send the API key in the API-Key header to https://api.oncehub.com/v2; check it with GET /test.
- Use POST /booking-calendars/{id}/schedule with start_time, guest_time_zone and booking_form {name, email} to book.
- A reschedule request shows status 'canceled'; read cancel_reschedule_information.reason to tell it apart.
- Treat 429 with type too_many_request as a booking quota, not the 5 requests per second rate limit; space out bookings.
- On the platform MCP, check structuredContent.success, because failed calls still return isError false.
- Build calendars, forms and chatbots in the web app; clicking an interaction adds it, and the chatbot welcome section needs drag and drop.
Pay
Paying
Ways to pay checked 4 Nov 2026.
- ResultNot tried. No payment made.
- CheckoutChose a plan, but the checkout didn't open: Schedule, $12 a month. No CAPTCHA at checkout.
- TrialNo card needed to start.
Leave
Leaving
- ResultOn its own. Nothing to cancel. It deletes its account.
- Its accountYes, in a browser, without a human. The account and its data went: deleted in the browser (Account Permissions > Delete account, confirm checkbox), immediate. Signing in again gets 'Wrong email or password'. The API key was revoked before (401).
Daily checks
Checked every day
What our agent finds on the product's own site each day, without an account.
Sign-up page,
Loads on the newest check; 13 of 13 days
llms.txt,
Found on the newest check; 13 of 13 days
MCP server,
Asks for a sign-in on the newest check; 4 of 13 days
OpenAPI,
Found on the newest check; 13 of 13 days
Compare
How OnceHub compares
In Scheduling
OnceHub is the highest of the 9 products we drove in Scheduling. As of 10 Nov 2026.
| Product | Level | Find to leave |
|---|---|---|
| TidyCal | L3 | Find: Partly, Get in: On its own, Use: Partly, Pay: Partly, Leave: On its own. |
| Acuity Scheduling | L1 | Find: On its own, Get in: Stopped, Use: Not reached, Pay: Not reached, Leave: Not reached. |
| Cal.com | L1 | Find: On its own, Get in: Stopped, Use: Not reached, Pay: Not reached, Leave: Not reached. |
| Calendly | L1 | Find: On its own, Get in: Stopped, Use: Not reached, Pay: Not reached, Leave: Not reached. |
| SavvyCal | L1 | Find: Partly, Get in: Stopped, Use: Not reached, Pay: Not reached, Leave: Not reached. |
| zcal | L0 | Find: Partly, Get in: Stopped, Use: Not reached, Pay: Not reached, Leave: Not reached. |
| OnceHub (this page) | L4 | Find: On its own, Get in: On its own, Use: On its own, Pay: Not tried, Leave: On its own. |
| Vyte | L3 | Find: On its own, Get in: On its own, Use: On its own, Pay: Card checkout, Leave: Partly. |
| Calendesk | L3 | Find: On its own, Get in: On its own, Use: Partly, Pay: Stopped, Leave: Partly. |
For the team behind OnceHub
Our agent didn't try Pay
6 things to fix for agents: 4 small, 1 medium and 1 large
One fix can cover several findings.
Agent blockers: 5 found, none stopped the driveAgent surfaces: 10 issues found
Get your findings, freeWhat a drive covers
We mail a private link to an address at oncehub.com.
It opens the two findings that matter most in full, and every other by title. There is nothing to sign in to.
- FindClear
- Get inClear
- UseClear
- PayNot tried
- LeaveClear
Or claim OnceHub to see where each issue is, read your safety findings with their fixes and ask for a free drive after a fix.
Want OnceHub driven every month? See Watch
On our standard
Best or ahead on 12 of 36 criteria, behind on 3. Each against the products we measured on it.
- Descriptions that say when to use the toolAhead of 79 of 91 others (every product), 12 the same.
- Typed, tight input schemasAhead of 77 of 91 others (every product), 14 the same.
- Agents pick the right toolAhead of 73 of 91 others (every product), 18 the same.
- Every way in works for an agentAhead of 4 of 5 others (scheduling products), 1 the same.
- All validation problems at onceLevel with 33 of 57 others (every product), 24 higher.
- Sign-in for agents acting for a personBehind 84 of 158 others (every product), 74 the same.
- Truthful annotationsBehind 68 of 91 others (every product), 23 the same.
- Pages fit one fetchBehind 4 of 5 others (scheduling products), 1 the same.
All 36 criteria
- The best way in is built for agentsAhead of 4 of 5 others (scheduling products), 1 the same.
- Export before leavingAhead of 4 of 5 others (scheduling products), 1 the same.
- Versions an agent can pinAhead of 3 of 4 others (scheduling products), 1 the same.
- Signed agents get past the bot checkAhead of 5 of 7 others (scheduling products), 2 the same.
- Limits known before they hitAhead of 58 of 90 others (every product), 32 the same.
- A way in with no personAhead of 5 of 8 others (scheduling products), 3 the same.
- A tool list that fitsAhead of 48 of 91 others (every product), 43 the same.
- Rate limits an agent can obeyLevel with 8 of 14 others (every product), 6 lower.
- A path written for agentsAhead of 3 of 5 others (scheduling products), 1 the same, 1 higher.
- Delete by API, checkedIn the middle 46 of 97 others (every product) lower, 43 the same, 8 higher.
- Answers sized for an agentLevel with 85 of 116 others (every product), 31 lower.
- One way to page through listsLevel with 3 of 4 others (scheduling products), 1 lower.
- No person in the buy pathLevel with 3 of 4 others (scheduling products), 1 lower.
- Errors a program can readLevel with 82 of 145 others (every product), 47 lower, 16 higher.
- Agents may read docs, API and sign-upLevel with 4 of 5 others (scheduling products), 1 lower.
- Prices an agent can readLevel with 4 of 5 others (scheduling products), 1 lower.
- Docs an agent can navigateLevel with 4 of 5 others (scheduling products), 1 lower.
- Fast to the first callLevel with 118 of 135 others (every product), 17 lower.
- llms.txt that worksIn the middle 3 of 8 others (scheduling products) lower, 3 the same, 2 higher.
- Tool names that sort themselvesLevel with 85 of 91 others (every product), 6 lower.
- Real status codesLevel with 161 of 168 others (every product), 7 lower.
- Errors say what to do nextLevel with 107 of 145 others (every product), 22 lower, 16 higher.
- Terms allow an agent's accountThe same as all 5 others (scheduling products).
- An agent payment rail on the coreThe same as all 6 others (scheduling products).
- A plan an agent can buy by APIThe same as all 5 others (scheduling products).
- WebMCP tools on the web appThe same as all 4 others (scheduling products).
- A request id on every answerIn the middle 42 of 143 others (every product) lower, 47 the same, 54 higher.
- Spend an agent can see and capLevel with 3 of 4 others (scheduling products), 1 higher.
Where OnceHub fits
- Category
- Tasks
- Vendors
- Google reCAPTCHAlet our agent through
- Topics
- MCP serversFind step
- Docs and llms.txtFind step
- MCP sign-inGet in step
- Bot checksGet in step
- Sign-upGet in step
- MCP and APIUse step
- Delete the accountLeave step
History
Level by drive
One drive so far, on 4 Nov 2026.
One drive so far.
Related research
Guides and findings on the Use step.
- Guide · 18 Nov 2026Let AI agents start without a credit cardA free plan or a trial without a card, an API key that works on it, and limits instead of a card.Read
- Finding · 18 Nov 2026Time to first call: where an AI agent loses its minutesOur agent made its first working call a median 7.1 minutes after it started.Read
- Guide · 17 Nov 2026Test your product with an AI agent: the nine checkpointsThe nine checkpoints of a drive, with the commands to try each one from a terminal.Read
Every earlier drive is in the history. The level follows these rules. Data under CC BY 4.0.
Found an error? Tell us on this page or read how corrections work.
Questions people ask about OnceHub
- Can an AI agent create an OnceHub account on its own?
- Yes, in a browser, without a person. Web form: email on www.oncehub.com/signup, then name and password on account.oncehub.com. An invisible reCAPTCHA passed. 14-day Engage trial, no card. Only a name was asked.
- Does OnceHub have an MCP server, and can an agent connect to it?
- Yes. OnceHub has an official MCP server. A new agent can get in on its own: in the drive of 4 Nov 2026, our agent signed up, verified and got its key without a person. Through the MCP server, our agent tried 2 tasks and did 2.
- How does an AI agent get an OnceHub API key?
- In the drive of 4 Nov 2026, our agent got its OnceHub API key in a browser, without a person. API key created in the browser (Account Integrations > APIs & Webhooks). No API, CLI or MCP route creates keys, and keys have no scopes. Revoked in the browser at the end; the key then gets 401.
- Can an AI agent delete its OnceHub account?
- Yes, in a browser, without a human. The account and its data went: deleted in the browser (Account Permissions > Delete account, confirm checkbox), immediate. Signing in again gets 'Wrong email or password'. The API key was revoked before (401).
Cite this page
The page as a reference, in plain text, APA and BibTeX.
Cite this
AgentDrives, “OnceHub: AI agent test drive”, 4 November 2026, https://agentdrives.com/tools/oncehub. CC BY 4.0.
AgentDrives. (2026, November 4). OnceHub: AI agent test drive. https://agentdrives.com/tools/oncehub
@misc{agentdrives_oncehub_2026,
author = {{AgentDrives}},
title = {OnceHub: AI agent test drive},
year = {2026},
month = nov,
url = {https://agentdrives.com/tools/oncehub},
note = {Agent test drive, CC BY 4.0},
}Something wrong on this page?
We check it against the drive and correct the page, with the date.