Build & Automate Your Second Brain: Web Clipper, GitHub Actions, Foam for VSCode

In the age of information, knowledge is power. But sometimes, the sheer volume of information available can become overwhelming. There’s a constant struggle to capture, categorize, and efficiently utilize the myriad of resources floating around in the digital world. This is where the concept of a “second brain” comes into play.

Photo by Milad Fakurian on Unsplash

A second brain is a system of tools and methods that helps you manage information, streamline your workflow, and leverage the content you come across daily.

By automating my second brain, I have been able to turn chaotic information streams into structured knowledge. Let’s dive into how I have managed to do this.

1. Web Clipper: Archiving Web Content to GitHub

All too often I’ve seen myself with 100 tabs open, trying to keep track of all the information I came across during an average web hunt. This is where Web Clipper comes in handy.

Web Clipper is a tool that lets you effortlessly save web content to your preferred platform, such as GitHub. I’ve used it to archive everything from articles to snippets of code. From there on out, I can easily search and access the information whenever I need it.

How It Works:

1. I come across a valuable piece of information on the web.
2. Using the Web Clipper extension, I clip it and save it directly to my GitHub repository.

🎉 Now, it’s permanently archived, categorized, and easily accessible for future reference.

You can find the open-source code for Web Clipper on GitHub and in the Chrome Web Store.

2. GitHub Actions: Automating Content

With GitHub Actions, I have been able to automate various content processes, like sending articles to my Kindle or archiving content as Markdown files. It’s like having a virtual assistant that works around the clock in the background.

Some Automation Ideas:

• Converting web articles into Kindle-friendly formats and sending them directly to your device.
• Archiving content as Markdown files.
• Summarizing content using language models — e.g. I can have content summarized using ChatGPT, extracting only the most critical information and why it could be relevant to me.

The possibilities are endless and already opened up a whole new world of possibilities for me since now I can read articles on my Kindle when I’m offline — or read the summarized version on my phone when I’m on the go.

The GitHub Actions management interface: Let’s you keep track of actions active in your repository and whether they succeeded

3. Writing GitHub Actions using ChatGPT: Leveraging AI

GitHub Actions are written in YAML, which is a markup language. This means that you can use any programming language to write them. Luckily enough, besides many GitHub actions that are already available, you can also write your own.

But don’t worry, you don’t need to be a programmer to do this. You can use a language model like ChatGPT to generate the code for you. All you need to do is provide a few examples of what you want the action to do, it will generate the code for you and after a few tweaks, you’re good to go.

Generating GitHub Action workflows using ChatGPT: The large language model gives good initial inspiration

The .github/workflows folder contains all workflows associated to a repository. After some debugging, I organized them based on actions happening in the repository

• append_dates.yml: Appends a new date & a daily briefing to my journal notebook
• assign_issues.yml: Runs when an article is created, assigns it to me & summarises its content
• handle_labels.yml: Runs when labels are changed to save an article as MD or send it to my Kindle device (e.g. “kindle” label triggers sending the article to my Kindle)

The .github/workflows folder contains all workflows associated to a repository

4. GitHub as a personal CMS

Organizing content using GitHub Issues / Project view

The project view shows articles stored in my personal GitHub repository which I can now keep easily keep track of

Article view

🪄Automated content summary

I created a GitHub action, which for every article generates a tailored summary that I can read on the go. Here you can find the prompt behind the automation:

Prompt: “You are a helpful assistant summarizing articles to maximum 200 words (explain it like I am five (ELI5))
Sell this article to me in a fun way (Headline: ## ✨ Summary), summarise. Add up to five fitting topic tags (1–3 very generic, up to 2 more specific) about the summarized text to the end of the summary in the format: ## Tags (new line) #tag1 #tag2 #tag3… Explain if and how this could be relevant as a UX designer with a concrete (possibly humorous) example (Headline: ## Personal Relevance) — in MD format

Besides a summary, the used prompt delivers tags. Why tags, you might ask? Using tags, the knowledge management system can make sense of the relation between different articles.

5. Foam as a Second Brain: Organizing and Linking

Foam is a personal knowledge management system that turns a simple directory of Markdown files into a linked network. It allows me to:

• Organize my notes and content in a logical, interconnected way.
• Quickly search and find relevant information.
• Create visualizations of my knowledge graph, identifying patterns and relationships.

Foam’s graph network shows patterns & relationships between content

Conclusion

Automating my second brain has revolutionized the way I interact with information. By leveraging tools like Web Clipper, GitHub Actions, and Foam, I have created a personalized knowledge management system that saves time, reduces stress, and empowers me to make the most of the information age. These tools are not just for tech-savvy individuals; they are designed with user-friendliness in mind. So, why not give them a try? Start building your second brain today and unleash the power of automation!

I’m also just starting with this. What are your thoughts on automating your second brain? What would you like to automate next? Let me know in the comments below 🤓

What’s next? Connecting it to privateGPT, a private AI model

Keeping your notes in a structured way is great, but what if you could also use them to generate new content? This is where privateGPT could come in handy. It’s a private AI model that can generate text based on your notes. The exciting part about it is that it is powered by a local large-language model of your choice, so you don’t need to upload your notes to the cloud. I am still experimenting with it, but I can already see the potential of it. How would you connect your ‘second brain’ to privateGPT?