What If the Tool I Need Has No Connector?
I named three routes when a participant asked about a tool with no connector: a tool built for public pages, an API connection, and a request to the platform.
Look through the connector list first. On October 7, 2026 I said there were over 300 inside Gumloop. If your system isn't there, I described three routes: for public web pages, a tool built for AI agents to read them; for a system that has an API, a connection that can take about 15 minutes; and a request to the platform to add the system.
A participant asked how to connect through an API if the tool has no connector (32:34). It's a fair question for an operations or IT lead whose system never shows up on a list. Below are the three answers I gave, set in the order you'd try them.
What do you check first?
The connector list. A connector is what gives an AI agent access to a system such as calendar, email or Slack. I told the group to take their time with it:
From the recording, 30:55. There are over 300 connectors that you can use inside of Gumloop.
Open your AI agent's settings, find Connectors and select the plus icon next to Add, as shown on October 7, 2026 (25:50). For which connectors to start with, see Which Connectors Does an AI Agent Need?. For what a connector is, see What Is an AI Agent Connector?.
On my relationship to Gumloop, I said this:
From the recording, 33:22. I don't work for Gumloop. I found this tool over the summer through Lenny's podcast, and I started using it and have since made myself an ambassador for them. I've created a relationship with Gumloop. I am not an employee.
More on that answer: Do you work for Gumloop? (the answer from the session).
What if the website has no connector at all?
I used LinkedIn as the example, because I use it to prepare for meetings:
From the recording, 28:16. There are a couple of websites that don't have connectors.
From the recording, 28:25. LinkedIn doesn't share the MCP and API. I'm getting a little technical here, but that's simply how one system talks to another.
My answer was a tool built for AI agents to read public pages. I named Apify:
From the recording, 28:51. Apify is an agent native tool that fetches specific information from various platforms such as LinkedIn.
It has limits, and I named them in the same breath:
From the recording, 29:18. If it's a private page, it will not read it, and also comments for LinkedIn.
So this route is for information anyone can already see. For a plain definition of that kind of tool, see What Is MCP, in Nontechnical Words?.
What if the system has an API?
When the participant asked, I said it can be more complicated, and I put a time on it:
From the recording, 32:48. That can be a bit more complicated. That may take about 15 minutes.
I told the participant we would hang on to it until the end of the session (32:47). The recording doesn't show that conversation, so I can't describe the steps.
What if you ask the platform?
I also told the participant I'd take the system name to the Gumloop team. On October 7, 2026 I said what I expected to happen next:
From the recording, 33:05. If they can do, they'll pull it in for us.
From the recording, 33:08. They typically ship in the same day, meaning they will service that ticket right away.
I do the same thing with features. On October 7 I said I had put in a service ticket with the team to switch something so we could select our model up front (19:03). Gumloop's documentation pages link a Contact support form, and its guide on choosing a connector, an API or MCP server, or the browser an AI agent can use lists the options in that order: a connector if one covers the app, then an API or MCP server, then that browser for everything else. It also says there are 300+ built-in connectors. Checked on 2026-11-05 against Gumloop's guide to connectors, APIs and the browser. My lines above are from October 7, and the docs page is what it says today.
The routes, in the order you'd try them
- 01
Check the connector list for your system.
- 02
For public web pages, use a tool built for AI agents to read them.
- 03
For a system with an API, expect a more complicated connection that may take about 15 minutes.
- 04
If you still need it, ask the platform to add the system.
What goes wrong
Three limits, all named in the session.
A site that doesn't share its connection. As I put it, LinkedIn doesn't readily share that information (28:34).
A private page. The tool I named won't read it (29:18).
An API connection that's more complicated than a connector, and may take about 15 minutes (32:48).
The recording doesn't show an API connection being built, a request being sent to the platform, or the answer.
Where this comes from
This article grows from the session I taught live with Women Applying AI on October 7, 2026. The quoted passages are my own words, taken verbatim from the recording. The steps, tables and checklists around them were drafted by my team's AI agents from those words and my published pages, and checked against the recording before publishing. Platform facts that have a public source carry the dated line above, and screens change.
Build Your First AI Agent in One Hour: the full session with the clips, the prompts and the transcript.
Related reading: Which Connectors Does an AI Agent Need?, What Is an AI Agent Connector?, What Is MCP, in Nontechnical Words?.
Frequently asked questions
How many connectors are there?
On October 7, 2026 I said there were over 300 inside Gumloop (30:55).
What do I do if my tool is not on the list?
Three routes from the session: a tool built for AI agents to read public pages, an API connection, or a request to the platform.
Can an AI agent read a private LinkedIn page?
The tool I use for LinkedIn won't read a private page (29:18).
How long does an API connection take?
It may take about 15 minutes and can be a bit more complicated, as I told the participant (32:48).
Build something that actually runs your workflow.
A focused, free 60-minute live session with Jacqueline. You build alongside her, on your own real task, and leave with an AI agent that is already running.