AI is getting increasingly good at understanding codebases now thanks to large context windows. Here are my suggestions:

1. Do you have MAX mode enabled? this is crucial for large codebases because it uses large context windows. but is more expensive
2. **use reasoning models. This is a must.** (gemini 2.5 pro, claude thinking, chatGPT o3). I prefer o3 the most
3. Use agent mode. This lets the model recursively investigate by running queries, looking at the outputs, and running more queries
4. Keep resetting the chat. You're better of doing a hard restart instead of keeping on correcting cursor for the 10th time. If it doesnt work once, improve your prompt and do a hard restart
5. **Context is incredibly important**. Make sure to spend enough time doing a clear write up on what you know, what you need. What the bug is. What the ideal state should be. Make sure to use @PullRequest to tag the correct branch/PR that did not have the bug. The model cannot read minds and talk to it like its your coworker. I like to use chatGPT's voice dictate mode and brain dump everything in a 5 minute yap session. Itconverts my yap to text. I edit it and then add in any examples that cant be dictated
6. Before you code, you need to spend enough time priming the chat window with enough context about the problem and codebase. Almost like generating a PRD
7. Use @Web to let cursor search docs
8. I'm also fundamentally shifting how I generate docs. I try to make ensure docs I write are searchable by LLMs by making sure if something important is in an image, theres keywords to ensure it can be searched by text

<https://docs.cursor.com/context/@-symbols/@-git>

<https://docs.cursor.com/guides/working-with-context> (this is gold)

<https://x.com/gdb/status/1878489681702310392> (how to write a prompt)

Sai Shreyas Bhavanasi

You should absolutely be using AI as much as possible. In my opinion, if you're not learning to code for the first time using AI rarely hurts your growth 

In fact using AI has helped me learn 10x faster. So many times when I see the AI do stuff and I'm like wow I had no idea I could do X,Y,Z

In my opinion in this age it's incredibly important to know how to use AI to 10x your velocity for the simple reason that your time is limited, but the AI's time spent writing code is infinite. So knowing how to best use AI will pay off huge. I also think there's a fundamental shift people need to learn to how to best use AI. 

**The most common failure mode I see is people trying to get the AI to work for you instead of trying to change your workflow to work with what the AI is best at - for the simple reason that AI is getting better and better and time spent running AI has no upper bound**

---

Given that I still think it's important to use AI to 10x your learning. When the AI outputs something make sure to ask it, why did you do that? how does that work? how does it handle XYZ use case. Dont just take the output and run with it. Actually drill deeper.

By doing this I've accelerated my learning so much

---

I also found taking the courses on debugging/coding and accelerating that process with AI to be super helpful. A lot of the stuff in Taro debugging is your approach. It's not exactly "hard skills" but the mentality around it. I've used those lessons from Taro to change my mindset around how I debug and instead of manually doing it I just ask an AI to do it.

---

The caveat I'll say is a lot of it depends on the underlying LLM. I am not sure about the quality of Meta's LLMs but I am guessing its not as good as using cursor + sota models.

Here are some excellent guides on how to get started with using AI: <https://docs.cursor.com/guides/advanced/large-codebases>

I would use Gemini 2.5 flash. Free 1500 requests/day from google

The goto these days is windsurf. I'm getting a much higher code output from windsurf than cursor and works especially well on large codebases

testing

Starting at a new team or company? Or maybe you're just picking up a new tech stack? Go through these resources to learn how to ramp up fast.

[Taro Top 10] Effective Onboarding

Software engineers can have impact in many ways outside of coding, but make no mistake: Writing tons of very high-quality code quickly will always be useful. Follow the resources here to learn how to become a "code machine".