I think your question is too large. First of all, honey pots and rugpulls are very different.
While honey pots make money with "bad" hackers, rugpull projects steal honest people.
There are many method if you want to identify scams. In the same time, I would say there is no way to be 100% sure (how to technically define a honey pots or a rugpull ?)
Here are a some of the methods I would use to do it.
Many honey pots use exactly the same template. Try to identify the maximum you can. Then you just need to compare the smart contract to your database
AI method. You may train an AI model to detect if a smart contract looks like one or not.
Try to identify each specific address store in the smart contract and create as much scenario as you can and see if one make someone being able to transfer unexpected tokens.
Knows that Certik recently has audited a rug pull. So even good auditors are not 100% safe in their check.