wrote a lot about how to increase the performance of Claude Code. I wrote about some techniques I use whenever I use the tool. However, there are other areas where you can focus to find out more about Claude Code. Today, I will be talking about one of them: Making all information accessible in Claude Code.
I'm going to discuss how I make sure that almost everything I do is tracked and stored somewhere on my computer, inside some main folder, where all the information is contained. I, for example, maintain:
- All my marketing stuff, like LinkedIn posts, X posts, and webinars
- All the code I write
- The information I receive
And all the other information I come across during my working day. If it can't be stored locally on my computer, it should at least be accessible to Claude Code, for example, via the Notion API.
I was not sponsored by Claude Code in the writing of this article; I am simply an avid user of the product.
Why do you need to make information accessible?
Most developers know that it is a best practice to document everything they do. However, documentation can sometimes feel time-consuming and boring, and it doesn't happen all the time.
I would argue, however, that not logging much of what you do is time-consuming in the long run. You spend a lot of time trying to find information that isn't there or you have to find information again because you didn't save it ahead of time.
Getting in the habit of keeping track of all the information you encounter is very helpful, especially if you can easily access all of this information through an AI agent. In addition, if you are actively using an AI agent to complete many of your tasks, it will do better because you have given it more context with the scripts you have provided.
Simple setting:
You should save all your data because it saves time and makes your AI agents work efficiently.
How to make all information accessible in Claude Code
Making all the information accessible in Claude Code actually consists of two steps.
- Save all the information you come across.
- Save it in a folder that Claude Code can access.
I will have one section for each step to go deeper into how to do it in practice.
Save all the information you come across
This task sounds very simple, and in theory, it is very simple, because you just need to keep all the context somewhere on your computer. In practice, however, it becomes more difficult. The reason for this is that storing such information can be time-consuming, and in most cases, information is only transmitted orally and not in written form, so you must take concrete steps to preserve it.
I have two recommendations to make storing all the information easier. Number one is using a transcription tool like MacWhisper or Superwhisper. These tools make it very easy to press one button, talk to your computer, and write down quickly.
This lowers the limit of writing a lot of information to your computer, which is very important if you want to store a lot of information. In addition, this is very useful if you are promoting the Claude Code as well, and it saves me many hours every week. This week alone, it has already saved me four hours of writing.
The second is to keep the motto of keeping all information in the back of your head at all times. If you keep reminding yourself of the benefits of writing things down and making them available later, it will be easier to use this system. I, for example, have some task reminders that remind me to write down general information I've learned a few times a week.
In addition, you must give Claude Code access to all the pieces of information it needs access to. You can, for example, give it access to your Notion page where you store many of your documents. You can give it access to mail or Slack messages. Of course, you need to pay attention to security concerns about connecting to these apps.
If you want a safer option, you can download content as PDF or text files and save them locally, which, of course, reduces or completely eliminates many security concerns.
I personally download the most important pieces of information, such as the relevant customer profile, documentation for our CDK stack, marketing materials, and so on, to save them as local files and make them accessible to my agent in the future.
Access information
Now that you have saved all your information, you just need to access it. There are, of course, many tools out there that can access or search information folders and access external information through APIs.
Personally, I use Claude Code because it has an easy setup. You can easily access every piece of information that I store locally on my computer, and I've noticed that it works very well for navigating through my folders and finding the right files and information that I want to find.
Claude's code does not use vector search and instead uses bash commands to search in folders. Anthropic itself noted that early versions of the tool started with vector searches, but they found that searching files with bash commands gave higher quality output, albeit slower in most cases.
If you want to access faster, you can consider other options that target your folder and make it more quickly accessible by retrieving the improved generation.
Some examples of this are Warp Terminal, where you can point to code bases, and a good example of a tool I use when I need to find information quickly on my computer. If I know a piece of information is there, and I just need to find a location for it, I'll use Warp instead of Cloud Code because I know it can access the information faster.
Another option you have is to use the Cursor. It crawls and indexes your code bases and can access information quickly, just like Warp. There are many other great options out there that do the exact same thing.
What to think about when you save all the information
The first thing you need to think about when storing all your information is security concerns. Of course, while making all this information accessible is incredibly powerful and will make you and your writing agent more efficient, it's also dangerous because all the information is available locally on your computer, and sent to LLM, which has its own security concerns.
One thing you might think about is API keys. You may store multiple of your API keys locally. And you should try to avoid files containing these keys to be read and sent to an LLM inference provider such as Google, OpenAI, or Anthropic.
Another thing to pay attention to is permissions. When you run Code Claude, it will ask for permission to run commands to read certain folders.
You can enable this automatically by authorizing certain commands or simply by using malicious override permissions. This will naturally bypass all permissions prompts and make all files accessible to Claude Code without needing to ask for permission. Yes, this makes information retrieval more efficient because you don't need to interact with the AI as much, but you also run the risk of it accessing information it shouldn't.
Overall, it's up to you to ensure you're following security best practices while maintaining the efficiency gains you get by storing all your information locally and accessing it through a code agent.
The conclusion
In this article, I discussed how I keep all the information I encounter locally on my computer to make it available to my coding agent. This makes me very successful because I can easily download information that I have encountered before. And when my coding agent does the work, it has access to more context. This is very important if you want your coding agents to perform well and act according to your preferences. So I highly recommend trying to keep as much information as possible, making it easily accessible to you and your coding agents.
👉 My Free eBook and Webinar:
🚀 10x Your Engineering with LLMs (Free 3-Day Email Course)
📚 Get my free ebook Vision Language Models
💻 My webinar on Vision Language Models
👉 Find me on social media:
💌 Stack
🐦 X / Twitter
