Founder @ https://xitter.cf/t.co/4gxqy9ziRO Founder @ https://xitter.cf/t.co/BKQTflQ4ao
Joined July 2015
- Tweets18.2K
- Following577
- Followers19K
- Likes57.5K
Idk ultra still goes off in the weeds without a written plan.
Why do we still need plan mode in Oct 2026?
Terra 6.1 ultra might be the killer move
Almost as good as Sol. Way faster than Sol. Reasonable usage (doesn't drain your weekly limit instantly)
Feels like you're actually making progress instead of watching paint dry
Ivy Link got my watch website scraped hard by OpenAI / Anthropic.
So far no traffic from them but we will see if it shows up in AI results soon
Been working in low/med lately just so I have a finger on the pulse
Tried ultra/ultracode this past week and things were good but took way too long to get a result and I had no idea how it was built.
3 clicks today and many more pages indexed.
Google only quickly indexes pages it deems helpful, reliable, people-first content.
The Ivy Link agents did all of this.
They also built some calculators and comparison pages today. Totally unprompted.
First 24 hours we have some impressions.
But it looks like so far only the homepage has been indexed
This is way better than I thought it would ever be.
New Ivy Link demo site: watchpricetracker.com
I built the site with Claude and then Ivy Link took over everything.
Ivy Link started exploring live/sold listings for watches to build out a product list.
Then it scraped the data and pulled it into it's own isolated customer database (almost impossible from where it actually got the data, that's the secret sauce).
From the DB data it laid out the page structures for the website.
Created the skeleton and basic data.
Used that seed to do enrichment (using the enrichment agent, of course) - searching for more data for each watch model to feed questions about those watch models.
Automatically enriched the pages and pushed a pull request.
While doing that the technical seo agent completed a lighthouse SEO audit. Removed some heavy scripts / css loading and several other issues like open graph images (did I mention I have a design agent for things like that?)
Automatically pushed and merged a pull request.
The keyword research agent found some gems and isolated pages to target like "Cheapest Rolex".
Automatically made a page for it based on the data (amazing page honestly).
Made a whole bunch more pages like that.
Then, I made a Google Search Console account for the site this morning, connected it to Ivy Link, and it automatically added the sitemap.xml for me for indexing (thanks Ivy Link!).
I will be reporting on the actual SEO metrics through out the week. It will be interesting to see how quickly it gets indexed and how fast before our first click.
While I was typing this out it found another content opportunity and it's building that now 😅
This is WILD
This is way better than I thought it would ever be.
New Ivy Link demo site: watchpricetracker.com
I built the site with Claude and then Ivy Link took over everything.
Ivy Link started exploring live/sold listings for watches to build out a product list.
Then it scraped the data and pulled it into it's own isolated customer database (almost impossible from where it actually got the data, that's the secret sauce).
From the DB data it laid out the page structures for the website.
Created the skeleton and basic data.
Used that seed to do enrichment (using the enrichment agent, of course) - searching for more data for each watch model to feed questions about those watch models.
Automatically enriched the pages and pushed a pull request.
While doing that the technical seo agent completed a lighthouse SEO audit. Removed some heavy scripts / css loading and several other issues like open graph images (did I mention I have a design agent for things like that?)
Automatically pushed and merged a pull request.
The keyword research agent found some gems and isolated pages to target like "Cheapest Rolex".
Automatically made a page for it based on the data (amazing page honestly).
Made a whole bunch more pages like that.
Then, I made a Google Search Console account for the site this morning, connected it to Ivy Link, and it automatically added the sitemap.xml for me for indexing (thanks Ivy Link!).
I will be reporting on the actual SEO metrics through out the week. It will be interesting to see how quickly it gets indexed and how fast before our first click.
This is way better than I thought it would ever be.
New Ivy Link demo site: watchpricetracker.com
I built the site with Claude and then Ivy Link took over everything.
Ivy Link started exploring live/sold listings for watches to build out a product list.
Then it scraped the data and pulled it into it's own isolated customer database (almost impossible from where it actually got the data, that's the secret sauce).
From the DB data it laid out the page structures for the website.
Created the skeleton and basic data.
Used that seed to do enrichment (using the enrichment agent, of course) - searching for more data for each watch model to feed questions about those watch models.
Automatically enriched the pages and pushed a pull request.
While doing that the technical seo agent completed a lighthouse SEO audit. Removed some heavy scripts / css loading and several other issues like open graph images (did I mention I have a design agent for things like that?)
Automatically pushed and merged a pull request.
The keyword research agent found some gems and isolated pages to target like "Cheapest Rolex".
Automatically made a page for it based on the data (amazing page honestly).
Made a whole bunch more pages like that.
Then, I made a Google Search Console account for the site this morning, connected it to Ivy Link, and it automatically added the sitemap.xml for me for indexing (thanks Ivy Link!).
I will be reporting on the actual SEO metrics through out the week. It will be interesting to see how quickly it gets indexed and how fast before our first click.
Agents are gathering watch data
They have access to their own db and scraper tools. Static site is generated from the db data.
Watches it is.
Watch this space.
Let's play a game. I will start a brand new site and see if the ivyl.ink process really works for driving traffic.
Pick a niche:
21%GPU price index
24%Pokémon card price guide
21%Retired LEGO set values
34%Watch market index
29 votes • Final results