1) Do you have freelancing or other experience which has exposed you to multiple people with similar problems? For example, have you implemented the same "$()#% authentication system for a web application 20 times? Package up that one little piece of the puzzle into a scalable way to teach people to do it without needing to have you or someone similarly skilled on their team. Common form factors include ebooks. Sell the ebook. If you want to sell lots of the ebook, start by offering some free incentive to get people to trade you their email address, then send them email about $TOPIC for a while to make them trust you as an authority on it, then ask them to buy your thing. (Note: You don't have to be the world's leading expert on building authentication systems. You've just taken somebody's shilling to do it twenty times, which means that it is likely much, much cheaper for somebody to buy your thing and hand it to his junior developer than it is to pay a similarly-experienced engineer to do that part of the system.)
2) Failing that, talk to businesses. I could give you a vague fact pattern to ask about (a problem which is amenable to a solution with code) but people seem to get hung up on that so I'll give you something really specific: you are looking for a MS Excel spreadsheet which has ever been mailed by Bob to Cindy, had Cindy edit it, and mail it back to Bob. Every time this happens a SaaS app gets its wings. Now go out and find, in the actual physical world, ten firms which have that same darn spreadsheet. Offer to build them a software system which solves the business problem which that spreadsheet represents. Ask if they would pay (pick a number based on how big the firm is) $50 to $250 a month for it. If yes, ask them to commit to buying it when it is ready. If you get 5 commits, build it, sell to them (n.b. you'll lose some commits here), and then start trying to sell it in more scalable Internet-y ways.
3) Failing that, Bingo Card Creator, an app which does everything wrong in terms of business model and market selection, sold ~$1k a month something like 6 months after launch just because I got halfway decent at organic SEO. If you're willing to get good at one generalizable acquisition strategy (SEO, Facebook ads, AdWords, etc etc), with that wind behind you even turkeys can fly. (It often turns out that there are more lucrative options than pushing turkeys around.)
4) Find or create one proprietary data source which is not currently exposed to Google which answers a question that demonstrates commercial intent. Expose that proprietary data source to Google. The federal government has approximately 100,000 CSV files of interesting data which are not helpful to someone asking questions like (not a good example site, but a great example of there-is-an-answer-to-this-buried-in-a-free-CSV) "What is the median salary of mid-career dentists in Topeka, Kansas?"
How you monetize a site like that depends on the specifics of the niche and how savvy you are about it. Lead generation is very lucrative, if the question you're answering tends to suggest near-term commercial intent in something with a liquid market. If you want to make a lot less money in a braindead simple fashion, just slap adsense on it.
Also, I think I alluded to this above, but anybody capable of executing on any of these has ipso facto developed skills which can be employed by Real Companies with Real Budgets to make hundreds of thousands of dollars in marginal revenue. I know you said you want passive options but more active options, like selling consulting, are things that might well fit where you are in life a few years down the road.