Seth Rose
@sethrose.dev
🤖 Building AI agents 🚀 Open Source Dev
💻 Self-taught programmer 💛 Cancer survivor
📷 Founder @ http://MomentMend.com
💻 Self-taught programmer 💛 Cancer survivor
📷 Founder @ http://MomentMend.com
Say no once, you look difficult.
Say no ten times, you look disciplined.
Say no ten times, you look disciplined.
September 23, 2025 at 10:16 PM
Say no once, you look difficult.
Say no ten times, you look disciplined.
Say no ten times, you look disciplined.
Your focus blocks aren’t luxuries, they’re armor.
If you don’t guard them, clients will fill the space with their chaos.
What’s the one block on your calendar that never moves?
If you don’t guard them, clients will fill the space with their chaos.
What’s the one block on your calendar that never moves?
September 22, 2025 at 4:13 PM
Your focus blocks aren’t luxuries, they’re armor.
If you don’t guard them, clients will fill the space with their chaos.
What’s the one block on your calendar that never moves?
If you don’t guard them, clients will fill the space with their chaos.
What’s the one block on your calendar that never moves?
Flexibility isn’t free labor. Adapt on your terms, not theirs.
If clients want evenings, shift your mornings later.
Today, I'm allowing for a peaceful morning.
If clients want evenings, shift your mornings later.
Today, I'm allowing for a peaceful morning.
September 22, 2025 at 1:56 PM
Flexibility isn’t free labor. Adapt on your terms, not theirs.
If clients want evenings, shift your mornings later.
Today, I'm allowing for a peaceful morning.
If clients want evenings, shift your mornings later.
Today, I'm allowing for a peaceful morning.
Momentum is built on small daily victories.
The size of the win matters less than the chain.
- Write them down.
- Show your receipts.
- Keep the streak alive.
What’s the smallest win you tracked today?
The size of the win matters less than the chain.
- Write them down.
- Show your receipts.
- Keep the streak alive.
What’s the smallest win you tracked today?
September 22, 2025 at 1:31 PM
Momentum is built on small daily victories.
The size of the win matters less than the chain.
- Write them down.
- Show your receipts.
- Keep the streak alive.
What’s the smallest win you tracked today?
The size of the win matters less than the chain.
- Write them down.
- Show your receipts.
- Keep the streak alive.
What’s the smallest win you tracked today?
The most radical productivity hack isn’t a cold plunge or a 5 AM alarm.
It’s turning distractions off at a set schedule everyday.
Sleep is the real leverage.
Be honest 😅 What time do you actually crash?
It’s turning distractions off at a set schedule everyday.
Sleep is the real leverage.
Be honest 😅 What time do you actually crash?
September 22, 2025 at 3:34 AM
The most radical productivity hack isn’t a cold plunge or a 5 AM alarm.
It’s turning distractions off at a set schedule everyday.
Sleep is the real leverage.
Be honest 😅 What time do you actually crash?
It’s turning distractions off at a set schedule everyday.
Sleep is the real leverage.
Be honest 😅 What time do you actually crash?
Boundaries aren’t walls, they’re safety nets.
They protect your time, your family, and your work by keeping focus where it matters most in each moment.
What’s harder for you, holding the line at home or at work (or both)?
They protect your time, your family, and your work by keeping focus where it matters most in each moment.
What’s harder for you, holding the line at home or at work (or both)?
September 21, 2025 at 11:03 PM
Boundaries aren’t walls, they’re safety nets.
They protect your time, your family, and your work by keeping focus where it matters most in each moment.
What’s harder for you, holding the line at home or at work (or both)?
They protect your time, your family, and your work by keeping focus where it matters most in each moment.
What’s harder for you, holding the line at home or at work (or both)?
Now I’m curious 🤔
If you had to hand off just one task to an AI agent…
📅 scheduling
📧 email drafting
🔍 research
📝 summarization
♻️ repetitive edits
Which one would you offload first if you could only pick one?
If you had to hand off just one task to an AI agent…
📅 scheduling
📧 email drafting
🔍 research
📝 summarization
♻️ repetitive edits
Which one would you offload first if you could only pick one?
September 18, 2025 at 3:06 PM
Now I’m curious 🤔
If you had to hand off just one task to an AI agent…
📅 scheduling
📧 email drafting
🔍 research
📝 summarization
♻️ repetitive edits
Which one would you offload first if you could only pick one?
If you had to hand off just one task to an AI agent…
📅 scheduling
📧 email drafting
🔍 research
📝 summarization
♻️ repetitive edits
Which one would you offload first if you could only pick one?
Summarization is my secret weapon.
Agents provide a clean TL;DR with action items. Skimming 10% of the content keeps me informed. It's the difference between drowning in information and managing it.
Agents turn the flood into a drinkable bucket.
Agents provide a clean TL;DR with action items. Skimming 10% of the content keeps me informed. It's the difference between drowning in information and managing it.
Agents turn the flood into a drinkable bucket.
September 18, 2025 at 3:06 PM
Summarization is my secret weapon.
Agents provide a clean TL;DR with action items. Skimming 10% of the content keeps me informed. It's the difference between drowning in information and managing it.
Agents turn the flood into a drinkable bucket.
Agents provide a clean TL;DR with action items. Skimming 10% of the content keeps me informed. It's the difference between drowning in information and managing it.
Agents turn the flood into a drinkable bucket.
The first job I give an agent is scheduling. You send an email that says, “When are you free?” and the back-and-forth begins.
An agent can check your calendar, check theirs, propose slots, confirm, and send the invite. It’s the most obvious time sink that people still tolerate.
An agent can check your calendar, check theirs, propose slots, confirm, and send the invite. It’s the most obvious time sink that people still tolerate.
September 18, 2025 at 3:06 PM
The first job I give an agent is scheduling. You send an email that says, “When are you free?” and the back-and-forth begins.
An agent can check your calendar, check theirs, propose slots, confirm, and send the invite. It’s the most obvious time sink that people still tolerate.
An agent can check your calendar, check theirs, propose slots, confirm, and send the invite. It’s the most obvious time sink that people still tolerate.
Up +7% in just over a month. No trading, just holding + staking.
Sometimes boring works. 😎
Sometimes boring works. 😎
September 15, 2025 at 4:15 PM
Up +7% in just over a month. No trading, just holding + staking.
Sometimes boring works. 😎
Sometimes boring works. 😎
My win: spent the past week focusing on social media interactions.
Engagement’s already up 60%+ so looks like it’s working.
Engagement’s already up 60%+ so looks like it’s working.
September 15, 2025 at 1:31 PM
My win: spent the past week focusing on social media interactions.
Engagement’s already up 60%+ so looks like it’s working.
Engagement’s already up 60%+ so looks like it’s working.
Indie hacking moat = consistency. Not secrets.
What’s your tiny win today?
What’s your tiny win today?
September 15, 2025 at 1:31 PM
Indie hacking moat = consistency. Not secrets.
What’s your tiny win today?
What’s your tiny win today?
AI agents often fail at tool handoffs, not reasoning.
The 3 design rules that fix it?
1. Ensure context persists through interactions.
2. Expect failures as the norm and plan accordingly.
3. Allow agents to seek clarification before finalizing.
Which ones do you commonly see ignored?
The 3 design rules that fix it?
1. Ensure context persists through interactions.
2. Expect failures as the norm and plan accordingly.
3. Allow agents to seek clarification before finalizing.
Which ones do you commonly see ignored?
September 15, 2025 at 3:30 AM
AI agents often fail at tool handoffs, not reasoning.
The 3 design rules that fix it?
1. Ensure context persists through interactions.
2. Expect failures as the norm and plan accordingly.
3. Allow agents to seek clarification before finalizing.
Which ones do you commonly see ignored?
The 3 design rules that fix it?
1. Ensure context persists through interactions.
2. Expect failures as the norm and plan accordingly.
3. Allow agents to seek clarification before finalizing.
Which ones do you commonly see ignored?
These are the 5 things that quietly decide if an LLM tool scales or breaks.
Which one do you think teams ignore most? 👇
Which one do you think teams ignore most? 👇
September 15, 2025 at 2:00 AM
These are the 5 things that quietly decide if an LLM tool scales or breaks.
Which one do you think teams ignore most? 👇
Which one do you think teams ignore most? 👇
Before launching your AI agent, make sure to save these 5 rules to catch weak AI agents before launch.
These common pitfalls are the primary culprits behind most AI agent launch failures.
Which one do you think kills scale fastest? 🧵
These common pitfalls are the primary culprits behind most AI agent launch failures.
Which one do you think kills scale fastest? 🧵
September 14, 2025 at 4:45 AM
Before launching your AI agent, make sure to save these 5 rules to catch weak AI agents before launch.
These common pitfalls are the primary culprits behind most AI agent launch failures.
Which one do you think kills scale fastest? 🧵
These common pitfalls are the primary culprits behind most AI agent launch failures.
Which one do you think kills scale fastest? 🧵
Why am I the way that I am? lol
github.com/TheSethRose...
github.com/TheSethRose...
May 22, 2025 at 5:25 AM
Why am I the way that I am? lol
github.com/TheSethRose...
github.com/TheSethRose...
HOT TAKE: Github Copilot is actually not bad, with the proper setup.
May 21, 2025 at 12:37 AM
HOT TAKE: Github Copilot is actually not bad, with the proper setup.
building in public isn't about success stories; it's about showing the actual process
May 16, 2025 at 6:03 AM
building in public isn't about success stories; it's about showing the actual process
Update. Looks like it completed with a few errors.
I'm now having it double check itself to see if it can resolve it.
I'm now having it double check itself to see if it can resolve it.
May 13, 2025 at 7:28 PM
Update. Looks like it completed with a few errors.
I'm now having it double check itself to see if it can resolve it.
I'm now having it double check itself to see if it can resolve it.
Update. Seems to be working really well.
May 13, 2025 at 6:05 PM
Update. Seems to be working really well.
Whelp, let's see if Claude Sonnet 3.7 can help me back up my files successfully.
Follow me for the update. 🤞
Follow me for the update. 🤞
May 13, 2025 at 5:00 PM
Whelp, let's see if Claude Sonnet 3.7 can help me back up my files successfully.
Follow me for the update. 🤞
Follow me for the update. 🤞
Planning to reset my Macbook Pro either tomorrow or the next day.
Typically use it for Development (Typescript & Python), LLMs and models.
- Got any pre or post reset rituals you'd recommend?
- What essential software do you always load up on a fresh install?
Typically use it for Development (Typescript & Python), LLMs and models.
- Got any pre or post reset rituals you'd recommend?
- What essential software do you always load up on a fresh install?
May 13, 2025 at 3:36 AM
Planning to reset my Macbook Pro either tomorrow or the next day.
Typically use it for Development (Typescript & Python), LLMs and models.
- Got any pre or post reset rituals you'd recommend?
- What essential software do you always load up on a fresh install?
Typically use it for Development (Typescript & Python), LLMs and models.
- Got any pre or post reset rituals you'd recommend?
- What essential software do you always load up on a fresh install?
