Thoughts by human, co-written by AI

lets work on the analytics, lets start a worklist on it, fanout but make sure each subagent writes research or any typa artifact in the analytics worklist folder so that all the artifacts get saved and we can iterate on things. at some point we will write the adr doc and proceed with engineering. For now you can look at the: https://gkoreli.com/stats i feel like something is wrong

proceed with engineering as needed

think about the domain driven engineering, design principle and design patterns. Make sure to decompose and reuse code. You have freedom to rewrite some of the existing pieces. Use inheritance and compositions as needed, I love composition whenever it makes sense.

wait a second, i wanna make sure the old data is not lost correct? everything is properly migrated?

what do you mean? i feel like old data was quite meaningful do you really understand it all?

why are you so careful about the legacy? this is my own blog, i just want to make sure the old data doesn't get lost, we don't need to be 100% accurate, dont you agree?

do you know why are we doing any of this at all? whats the goal

write adr 16.1 with this goal and intent in mind, and try to answer all these now with the existing architecture and engineering

any adjacent big impact work? do you think we should write an article on how to engineer analytics system for the blog? and use our own blog as an example? Would it be an interesting read, or a trendy or meaningful work to write and share right after the https://gkoreli.com/does-llms-txt-work?

start a worklist and make sure to save all the subagent artifacts in the worklist

i feel like having a blog right now comparing what we had built and how we drastically improved, lessons learned and shared with others are really valuable, in 30 days we can have another article that will be a continuation, but we already have a lot of data from previous analytics implementation, and new approach, what it improves why, and why did we make a decision to spend all these effort, all the aspects and engineering advices, principles, architecture, mental model, vision and all that...

i always say to write from the active pain, ambiguity and all of that. Waiting 30 days to write like an expert is not transparency, its not sharing the pain, and its not what it takes to be fully transparent and be vulnerable, be relatable, you need to write from the open wound, we have lots of data, prior art and lots of artifacts from the worklist, we have lots of lessons to share about, and lots of unresolved tensions, we love writing about the unresolved tension. Update the skills as needed, because all of this is really important to me and seems like i have to re-iterate over and over.

btw these rules are not rigid, like we can write another article in 30 days where we have much more expertise and much more lessons to share, but at that time we are going to have new challenges and new wounds and new tension points, we will be vulnerable and relatable again at that point and then we can write in the past tense about the lessons and experiences, and in the present tense about the unresolved tension and pain, this is really important to know, i don't want to only write about the wounds just because i love it, we need to share our lessons and about what we learn and grow, but overall my point is that we shouldn't be rigid, we can write about past and present and future, its all about what matters when, we can decide based on the context what to write. I dont want us to be rigid all of a sudden and only write one particular style of articles.

its all about following: sometimes you write present tense, write about the open wound, unresolved tension, ambiguity, not knowing, about making calculated best guesses and judgment calls, bets and all that from past experience and present conditioning and beliefs. When writing about the past we reflect, analyze, learn and grow from past examples and lessons, and share them to others to help them grow alongside with us. Another big vision tenet why we write at all is to help ourselves grow, help others grow and build in public, we are building in public, inviting a third person eye into our lives, our tensions and ambiguity, we not only write about our pain points and ambiguities but our projects, what we are building and how, what vision we have, what tenets we have, what projects bets we are making, how we are planning to change the world, and we write about all of this in past tense, present and future tense, all of it makes sense based on what we are writing and based on the context. This is really important, save as a tenet somewhere, or save as a prompt for later article and/or update the skills or skill references as needed.

save part of these mental models somewhere, it will be a really nice article explaining what we write, how and why at some point in the future.'

where is the article? write it

I Built a Correct Counter for the Wrong Event - i like this but this kinda article will not bring readers to our blog... i think it needs to be more external, like very naive example would be: how to build the analytics with full transparency for a personal blog, but this is not a good title, i am just using as an example

the language and positioning matters a lot, imagine what are potential ways to come across this website, what would someone be looking for and comes across this article and they find it very meaningful and interesting to read it. Can you reverse engineer the title and SEO positioning? We shouldn't change the core article itself, but headers, titles and certain keywords matter a lot.

where is the researchFootprint and totalTokens for this article? i feel like we spent so much time, effort and tokens on this article, it would be nice to showcase it

in the article you can't tell how new referrer works, it only talks about the wrong mistakes, and nothing about in the present tense, whats up with that? We need to show comparisons and rationale, this is the wrong one how it used to be, this is how it looks now, this is why i made this decision, rationale and tenet, vision. Why don't we have any of this? Its impossible to share a lesson learned if we only talk about the mistakes, but do not explain the takeaways or how did we fix them and why. Only what is not meaningful, how and why are really important as well, capture these in the skills or agents.md and make necessary changes. Once finished commit and push, update the article too.