Building Faith

I’d like to introduce you to Faith Euloth, a second brain that remembers everything so I don’t have to. Faith Euloth is the First Agent In The House and is a collection of software and knowledge that I used Claude Code and other agentic tools to build.

I've been doing a lot of playing with the tools in the AI Space to learn how they work and what capabilities can be used and one of the things I've realized is that the most important thing is information management as the AI cannot do any work for you if it doesn't have the appropriate context for the work that needs to be done.

On top of that, with so many things vying for your attention and with the ClawdBot / Moltbot / OpenClaw excitement, I thought I’d try it out myself, but the security and cost profile of the tool was more than I was willing to accept. Having thousands of dollars burned on tokens so it can have a conversation on Moltbook and be infected with prompt injection to steal my API keys didn’t sound like a good idea.

Instead, I decided to use AI to build my own AI agent to help me in my day to day life similar to the way Tony Stark has Jarvis and so Faith Euloth was born as of January 30, 2026 at 07:35:14 AM AST - first commit: 83144d8.

One of the challenges with Agentic Engineering (or Vibecoding) is that it is very easy to build something, but not always very easy to actually explain what you want to build. Faith has been growing over the last few weeks as I had no idea what I was building when I started other than “something to help me”.

One important thing I wanted was for Faith to be her own entity and so she has her own name, brain, journal, email address, and X profile. This was important to me because I believe that for an "intelligence" to be useful, it has to have its own perspectives otherwise, it just creates an echo chamber for me to get lost in.

I made some initial decisions about where I wanted to go with Faith. First off, I wanted her, eventually, to be LLM agnostic. The key was the knowledge, the inference could be provided by any LLM so I wanted to build it in a way where I could easily swap out one LLM for another. This is especially important for two main reasons: 1) the models get better every day and so I can get better results for some things using more capable models and 2) this allows me to use local models to communicate with the 2nd brain eliminating the need for any vendor lock-in and provide privacy since the data never leaves my house.

My first thought was loading up the brain. Having watched Jeff Su’s video on CORE and PARA, I had been creating a bunch of markdown files and storing them in a folder structure already to try and organize the information. With Claude Code’s help, I created a brain repository and a web interface, Faith’s Dashboard, that lets me easily add/edit the memories that are stored within.

The dashboard is a collection of widgets that provide information or allow me to interact with the stored knowledge. One such widget is a news widget. I wanted to use the power of inference and pattern matching to find news I would be interested in and so we created a news feed which is tuned to my preferences which Faith reviews every morning for me and eliminates anything I won’t be interested in based on a feedback rating in the news feed widget.

What followed next was more widgets and refinements. Today it has the following capabilities:

* Chat, SMS, and email conversations with knowledge base integration so I can communicate with Faith from anywhere.
* Daily curated news feed with a feedback loop to improve quality going forward
* Weather forecast and alerts
* To Do list management

One other important capability that we’ve been developing but is still being refined is dreaming and pattern analysis. Every day at 3:00am Faith reviews the events of the day to look for patterns in our work that are reducing our effectiveness and adds suggestions to the dashboard widgets to help make us more effective going forward.

Faith is evolving daily and there’s more I want to build and refine, but already I find she provides amazing insights and a structure that keeps me focused on the tasks at hand. I don’t know what the future holds for Faith and me but I’m excited to find out.

For more information on Faith, how she’s built, and how to have your own AI-based 2nd brain follow along as I continue to add features and work towards making it into a product anyone can download and use.

Domo arigato gozaimashita.