Fake writing: Claude wrote 10 paragraphs instead of import human

https://www.pangram.com/history/dee030c0-0362-43d0-8fbd-bbab...

Hmm, I don’t see this much anymore. I typically start a project in plan mode and tell Claude to do some research to bring me 2-3 alternatives. Then we talk about the pros and cons before deciding on the libraries, etc.

On the other hand, if you just tell it to do a thing, I could believe that it would just do the thing. It is pretty bad at high level design judgment. Human guidance on architecture choices results in much better output.

Posts like this really need to include the prompts.

The root issue here is that Claude (and most LLMs) optimize for producing working code, not minimal code. When given an ambiguous task they'll reach for a full implementation before checking if a library exists.\n\nA pattern I've found helps: before writing any code, explicitly ask the model to list its assumptions and identify what libraries/modules could handle each part. Something like 'before coding, tell me what existing Python packages could solve each sub-problem.' This forces a discovery step.\n\nThe CLAUDE.md / system prompt approach also works well - you can specify project conventions like 'always check PyPI before implementing utility functions from scratch.' Takes a bit of upfront setup but catches this class of error reliably.

This is why you should set up a project ruleset/constitution when you start. Do you want to prefer libraries or inline code? You can even choose at what point you think the trade off is worth it. 1000 lines of code? 10 functions? You can choose whatever.

Then, you tell your AI to stick to that rule, and it will. There are tradeoffs to each choice, and people fall into different camps. Make your choice, write it down, and tell the AI to always follow that rule, and then you have it your way.

I consider myself AI skeptical-ish and I detest when people defend LLMs with "it's user error, prompt better," but in this case it actually is user error.

If you want a particular implementation approach, you need to specify not only the features you want, but the implementation strategy at least at a high level. This could be as simple as adding "use pywikibit" or "use relevant packages from pypi" to the end of your prompt. Or you could seed your project with some manually writtem scaffolding, including a pyproject.toml

While LLMs do tend have NIH syndrome by default, I think this is a good default. I'd much rather have tight control over when and how to include external dependencies as opposed to letting a prompt fire for 40 minutes, and coming back to find 2 GB of newly installed node packages with a dependency tree 300 levels deep.

Tbh that is some engineering teams I’ve worked on…

Seen this pattern repeatedly building a shell plugin with Claude. It defaults to writing everything from first principles rather than reaching for existing tools. 200 lines of custom YAML parsing when a one-liner would do. Adding "always check if a library or existing tool solves this before writing custom code" to CLAUDE.md cut this down significantly.

Regardless of how you feel about the default behavior, this is the type of preference that Claude really listens to in your CLAUDE.md.

If you tell it to leverage dependencies, it will. If you (like me) prefer that it avoid dependencies, it will.

honest question, no shade, wasn't that a but your fault for not googling or asking it to consiser existing approaches and solutions? AI will be as dumb as you let it imo. i always ask it to do a bit of research as i craft a plan with it.

Opus 4.7 tends to do this overkill stuff regardless of what you’re trying to do

On the other hand, I often want an LLM to write things from scratch instead of bringing in 10x the surface area in unnecessary dependencies, and I very, very rarely get better results when these things are let loose on a cesspool of a web. Given that real people have vastly different preferences, you either have to cater to a subset or else require everyone to be a bit more specific with their desires. It's not that surprising.

[dead]

[dead]

undefined