Pres Mihaylov
@pmihaylov.bsky.social
eng at Plain, ex-Uber, building stuff with AI and claude code
The stuff:
- http://claudecontrol.com
- http://tinyurl.com/ai-eng-with-go
- http://vibegest.com
The stuff:
- http://claudecontrol.com
- http://tinyurl.com/ai-eng-with-go
- http://vibegest.com
Just added a heavily requested feature to vibegest.com, prob the most requested one - keeping track of twitter posts so you don't miss out.
It's quite demanding to doom scroll through the "For you" tab every day until you uncover the gems hidden in there.
It's quite demanding to doom scroll through the "For you" tab every day until you uncover the gems hidden in there.
December 20, 2025 at 4:10 PM
Just added a heavily requested feature to vibegest.com, prob the most requested one - keeping track of twitter posts so you don't miss out.
It's quite demanding to doom scroll through the "For you" tab every day until you uncover the gems hidden in there.
It's quite demanding to doom scroll through the "For you" tab every day until you uncover the gems hidden in there.
I rebuilt vibegest.com from the ground up:
- Track top news from HN and the subreddits you like
- Get the gist with a quick skim with article summaries
- Track people on who left X
- Track launches on YC, ProductHunt and more
Sign up and get your own vibegest.
- Track top news from HN and the subreddits you like
- Get the gist with a quick skim with article summaries
- Track people on who left X
- Track launches on YC, ProductHunt and more
Sign up and get your own vibegest.
December 19, 2025 at 10:30 AM
I rebuilt vibegest.com from the ground up:
- Track top news from HN and the subreddits you like
- Get the gist with a quick skim with article summaries
- Track people on who left X
- Track launches on YC, ProductHunt and more
Sign up and get your own vibegest.
- Track top news from HN and the subreddits you like
- Get the gist with a quick skim with article summaries
- Track people on who left X
- Track launches on YC, ProductHunt and more
Sign up and get your own vibegest.
It's been a while since I've written stuff on my blog and lots has happened since.
Here's an attempt at catching you all up on what I've been up to lately!
pmihaylov.com/what-ive-be...
Here's an attempt at catching you all up on what I've been up to lately!
pmihaylov.com/what-ive-be...
What I've been working on - pmihaylov.com
An overview of all my recent projects from the past few months.
pmihaylov.com
December 16, 2025 at 1:57 PM
It's been a while since I've written stuff on my blog and lots has happened since.
Here's an attempt at catching you all up on what I've been up to lately!
pmihaylov.com/what-ive-be...
Here's an attempt at catching you all up on what I've been up to lately!
pmihaylov.com/what-ive-be...
I tried rebuilding Vibe Jest to be fully agentic, driven by Claude Code instead of any code at all. So far, I haven't had any luck in making it perform well. I managed to get it to a point where it can send me a newsletter email, but the data it curated is not good at all.
October 21, 2025 at 3:59 PM
I tried rebuilding Vibe Jest to be fully agentic, driven by Claude Code instead of any code at all. So far, I haven't had any luck in making it perform well. I managed to get it to a point where it can send me a newsletter email, but the data it curated is not good at all.
I'm really excited about the new Claude Skills release because it is a significant UX improvement in how agents are built. Suddenly, anyone who can speak English can build an agent. No coding required.
In fact, you can speak to your agent and let it build itself.
In fact, you can speak to your agent and let it build itself.
October 21, 2025 at 2:59 PM
I'm really excited about the new Claude Skills release because it is a significant UX improvement in how agents are built. Suddenly, anyone who can speak English can build an agent. No coding required.
In fact, you can speak to your agent and let it build itself.
In fact, you can speak to your agent and let it build itself.
I stumbled upon a surprisingly good productivity hack. I set up an agent to run on a cron job, pick Linear tickets at random, and generate a PR for me while I sleep. In the morning, I review it and merge.
AI is literally writing code for me while I sleep
AI is literally writing code for me while I sleep
October 21, 2025 at 1:56 PM
I stumbled upon a surprisingly good productivity hack. I set up an agent to run on a cron job, pick Linear tickets at random, and generate a PR for me while I sleep. In the morning, I review it and merge.
AI is literally writing code for me while I sleep
AI is literally writing code for me while I sleep
I had to do some evals on an AI agent I built, and turns out Claude Code is very good at running evals. I just told it to run itself with a prompt continuously, find any bugs or issues in the underlying tools which is preventing the agent from performing well, and fix them.
October 20, 2025 at 3:57 PM
I had to do some evals on an AI agent I built, and turns out Claude Code is very good at running evals. I just told it to run itself with a prompt continuously, find any bugs or issues in the underlying tools which is preventing the agent from performing well, and fix them.
I've been playing with building AI agents off of various APIs.
What I found is that GraphQL performs significantly better than REST because GraphQL allows you to return the normalised data, whereas REST typically returns IDs and you need to enrich them separately.
What I found is that GraphQL performs significantly better than REST because GraphQL allows you to return the normalised data, whereas REST typically returns IDs and you need to enrich them separately.
October 20, 2025 at 3:02 PM
I've been playing with building AI agents off of various APIs.
What I found is that GraphQL performs significantly better than REST because GraphQL allows you to return the normalised data, whereas REST typically returns IDs and you need to enrich them separately.
What I found is that GraphQL performs significantly better than REST because GraphQL allows you to return the normalised data, whereas REST typically returns IDs and you need to enrich them separately.
Claude skills made MCP obsolete for me. I was able to build a Linear project manager skill in ~10min by asking claude to browse their docs + claude skills docs and build a new skill for itself.
I no longer need to use the Linear MCP for that.
I no longer need to use the Linear MCP for that.
October 20, 2025 at 1:57 PM
Claude skills made MCP obsolete for me. I was able to build a Linear project manager skill in ~10min by asking claude to browse their docs + claude skills docs and build a new skill for itself.
I no longer need to use the Linear MCP for that.
I no longer need to use the Linear MCP for that.
I just deployed a claude control agent on my server and it can now serve as my devops agent - anytime I need something done on my server, I can just ask it via Slack.
Eg here I wanted to stop a specific docker container for a couple mins and I didn't have to ssh to do it!
Eg here I wanted to stop a specific docker container for a couple mins and I didn't have to ssh to do it!
October 13, 2025 at 2:03 PM
I just deployed a claude control agent on my server and it can now serve as my devops agent - anytime I need something done on my server, I can just ask it via Slack.
Eg here I wanted to stop a specific docker container for a couple mins and I didn't have to ssh to do it!
Eg here I wanted to stop a specific docker container for a couple mins and I didn't have to ssh to do it!
Today I asked Claude Control to analyse the system I've deployed and summarise what it does. I couldn't have said it better
October 13, 2025 at 12:56 PM
Today I asked Claude Control to analyse the system I've deployed and summarise what it does. I couldn't have said it better
Maintaining docs is tedious and keeping CLAUDE md up to date is part of that. However, this is critical so your coding agent can improve over time for you and your whole team.
Fortunately, this is trivial to do when you can instruct claude to self-document its mistakes.
Fortunately, this is trivial to do when you can instruct claude to self-document its mistakes.
October 10, 2025 at 1:03 PM
Maintaining docs is tedious and keeping CLAUDE md up to date is part of that. However, this is critical so your coding agent can improve over time for you and your whole team.
Fortunately, this is trivial to do when you can instruct claude to self-document its mistakes.
Fortunately, this is trivial to do when you can instruct claude to self-document its mistakes.
Small tip to ensure your claude kicks off a task with the right context - ask it to study the changes you're building on top of and ask it to summarise them for you.
It forces it to gather the necessary context to pick off from where it left off without too much context rot.
It forces it to gather the necessary context to pick off from where it left off without too much context rot.
October 9, 2025 at 4:04 PM
Small tip to ensure your claude kicks off a task with the right context - ask it to study the changes you're building on top of and ask it to summarise them for you.
It forces it to gather the necessary context to pick off from where it left off without too much context rot.
It forces it to gather the necessary context to pick off from where it left off without too much context rot.
Claude Code has released a bunch of new features recently, but the one I'm most excited about is the ability to selectively enable/disable MCP servers.
This is a Big Deal because it enables you to have many MCP tools at stand-by without polluting your context
This is a Big Deal because it enables you to have many MCP tools at stand-by without polluting your context
October 9, 2025 at 3:00 PM
Claude Code has released a bunch of new features recently, but the one I'm most excited about is the ability to selectively enable/disable MCP servers.
This is a Big Deal because it enables you to have many MCP tools at stand-by without polluting your context
This is a Big Deal because it enables you to have many MCP tools at stand-by without polluting your context
Quick tip for all the vim nerds using claude code - Press Ctrl+G to write the prompt in vim!
October 9, 2025 at 2:01 PM
Quick tip for all the vim nerds using claude code - Press Ctrl+G to write the prompt in vim!
Claude Control is now far more reliable than it used to be - it can continue an ongoing session more than 24h since it started.
Making this work was surprisingly challenging!
Making this work was surprisingly challenging!
October 9, 2025 at 1:01 PM
Claude Control is now far more reliable than it used to be - it can continue an ongoing session more than 24h since it started.
Making this work was surprisingly challenging!
Making this work was surprisingly challenging!
Checking in all your claude artifacts like CLAUDE md, subagents, etc into the repo turned out to not just help me because claude can reuse these tools out of the box.
Turns out it helped a random contributor make a meaningful change from scratch without any setup!
Turns out it helped a random contributor make a meaningful change from scratch without any setup!
October 8, 2025 at 4:03 PM
Checking in all your claude artifacts like CLAUDE md, subagents, etc into the repo turned out to not just help me because claude can reuse these tools out of the box.
Turns out it helped a random contributor make a meaningful change from scratch without any setup!
Turns out it helped a random contributor make a meaningful change from scratch without any setup!
When I'm coding manually, I like to name my variables in a bit more cryptic way as a human watermark so they all know a human was here
October 8, 2025 at 3:01 PM
When I'm coding manually, I like to name my variables in a bit more cryptic way as a human watermark so they all know a human was here
Today I had to deal with a very nasty, intricate bug which required manual coding and intervention.
This happens quite rarely with claude code these days as most of the mundane work I've automated, which leaves my attention for these more challenging tasks instead.
This happens quite rarely with claude code these days as most of the mundane work I've automated, which leaves my attention for these more challenging tasks instead.
October 8, 2025 at 1:56 PM
Today I had to deal with a very nasty, intricate bug which required manual coding and intervention.
This happens quite rarely with claude code these days as most of the mundane work I've automated, which leaves my attention for these more challenging tasks instead.
This happens quite rarely with claude code these days as most of the mundane work I've automated, which leaves my attention for these more challenging tasks instead.
Very cool quality of life improvement for Claude Control I recently shipped - the agent is now made quite reliable to the point where it can continue working on a job even 24h later!
October 8, 2025 at 1:04 PM
Very cool quality of life improvement for Claude Control I recently shipped - the agent is now made quite reliable to the point where it can continue working on a job even 24h later!
Quick quality of life improvement for Claude Control - you can now start new jobs off of finished ones in the middle of a slack/discord thread 👌
October 6, 2025 at 4:04 PM
Quick quality of life improvement for Claude Control - you can now start new jobs off of finished ones in the middle of a slack/discord thread 👌
Looks like Anthropic starts counting the 5h window when you first use claude vs it being a constant fixed & rolling window.
So a tip to maximise usage is to trigger something quick like "hey Claude" the moment you wake up so the clock starts ticking during your morning downtime
So a tip to maximise usage is to trigger something quick like "hey Claude" the moment you wake up so the clock starts ticking during your morning downtime
October 6, 2025 at 3:02 PM
Looks like Anthropic starts counting the 5h window when you first use claude vs it being a constant fixed & rolling window.
So a tip to maximise usage is to trigger something quick like "hey Claude" the moment you wake up so the clock starts ticking during your morning downtime
So a tip to maximise usage is to trigger something quick like "hey Claude" the moment you wake up so the clock starts ticking during your morning downtime
One really nice thing about how Anthropic handles usage is that it doesn't cut you off halfway through a running session. This is great when you run background agents which are long-running by design.
October 6, 2025 at 1:58 PM
One really nice thing about how Anthropic handles usage is that it doesn't cut you off halfway through a running session. This is great when you run background agents which are long-running by design.
You can now deploy multiple claude code containers, effectively enabling you to have different repos routed to different discord/slack channels.
Now I have one channel for claudecontrol repo and one for ccagent
Now I have one channel for claudecontrol repo and one for ccagent
October 6, 2025 at 1:01 PM
You can now deploy multiple claude code containers, effectively enabling you to have different repos routed to different discord/slack channels.
Now I have one channel for claudecontrol repo and one for ccagent
Now I have one channel for claudecontrol repo and one for ccagent
First impressions of the new Claude Sonnet 4.5 model is that it is very, very fast to what the previous model did, and also it tends to spawn much more tasks in parallel, so it results in less idle time for me, which is great.
September 30, 2025 at 1:03 PM
First impressions of the new Claude Sonnet 4.5 model is that it is very, very fast to what the previous model did, and also it tends to spawn much more tasks in parallel, so it results in less idle time for me, which is great.