Start by writing down, in a sentence or two, what your idea is and which competences it is missing. Then put that sentence where people with those competences will see it: first the people you already know and the people they know, then communities, events and matching services. Whoever you find, work on a small piece together before you agree on anything big.
Decide what you are looking for
A co-founder shares the ownership and the risk, and is in it for years. A collaborator helps with a part of the work, for a while. An adviser gives opinions and introductions. If what you need first is help with one part for a few months, ask for a collaborator: it is a much smaller thing to ask, and easier to say yes to.
Name the competence, not the title. "Someone who can build the first version of the app" tells a stranger what you need and whether they could be the one. "A CTO" does not. The same goes for sales, design, law, hardware or the plumbing of a food business: say what has to be done that you cannot do.
Where to look
People you know. Tell friends, former colleagues and classmates the one-sentence version and ask who they would introduce you to. An introduction from someone both sides trust is worth more than a cold message.
Communities and events. Meetups, hackathons, courses and online groups for your field let you see how someone works before anyone mentions a partnership.
Matching services. Several services exist where people looking for a co-founder find each other. They differ in who they are for and what they cost; we compare the main ones in co-founder platforms compared.
Bay of Ideas. You publish the idea, name the competences you lack, and members who have them can find it and offer to help. The idea has a public side that anyone can read and a private side only the people you accept can see. It is free, and still small, so treat it as one place among several.
How to approach someone
Say what the idea is in a sentence, which part you want help with, how much time you are asking for, and what you can offer in return. Do not ask "do you want to be my co-founder?" in the first message. Ask for a conversation, or for help with one concrete thing. If the idea is sensitive, share the outline and keep the details for later; see where and how to share an idea safely.
Test the match before you commit
People are easier to judge by what they do than by what they say. Agree on a small task with an end date, such as a prototype, ten customer conversations or a first draft of the plan, and see how it feels to work together: do they deliver, answer messages, disagree in a useful way?
Before anyone puts in serious time or money, talk through the questions that tend to break partnerships:
How many hours a week each of you can give, and for how long.
Who decides what, and what happens when you disagree.
Who owns what share, and what happens to it if one of you leaves.
Whether anyone is paid, and where the first money comes from.
Write the answers down. If the venture is more than a hobby, have a lawyer draw up a proper agreement; a few paragraphs in an e-mail will not hold when the stakes are real.
Signs to slow down
The person is enthusiastic but cannot say how many hours they can give.
You have not seen any of their work, and they have not asked to see yours.
They want ownership before they have done anything, or want to avoid writing anything down.
You keep avoiding a difficult subject because the conversation is going well. It will come back.
It is fine to part after a trial and stay friendly. Finding the right person can take longer than you hope, and a good "no" early costs far less than a bad partnership later.