Vy - Cross platform AI agent — AI agent that uses your computer, cross platform, no APIs
AI agent that uses your computer, cross platform, no APIs
Promoted
Maker
📌
👋 Hey Product Hunt! I’m the maker behind Quercle.
# What inspired me to build this? I was building LLM agents and needed a simple web fetch + web search tool (similar to what Claude Code has). Everything I tried had issues: some returned raw HTML, others produced messy markdown full of irrelevant content, and many just broke on JavaScript-heavy sites. Cleaning and post-processing all of that logic became its own problem.
# What problem was I trying to solve? I wanted web tools that return exactly what an LLM should see - clean, relevant, structured data - without developers having to write custom scraping, parsing, or filtering logic for every site.
# How did the approach evolve? This originally started as a small part of another project. While working on it, I realized this piece was actually more feasible and clearly scoped than my original idea, so I decided to pivot and focus on it as a standalone product. The trade-off I chose was simplicity and LLM-ready output over raw speed: the data goes through an LLM layer before being returned, so it’s immediately usable in agents and workflows. It also works across JS-heavy sites and integrates easily with different frameworks.
# Why I’m launching: I’m mainly looking for feedback and testers. I want to know: - Does this solve a real problem for you? - Would you use it in production? - What’s missing?
---
It is free to try out! If you try it out and share feedback, I’m happy to give free credits - just reach out. Any thoughts (good or bad) are hugely valuable at this stage.
Have a question about Quercle? Ask it here and get a real answer.
Do you use Quercle?
Maker Comment
Maker
📌
👋 Hey Product Hunt! I’m the maker behind Quercle.
# What inspired me to build this? I was building LLM agents and needed a simple web fetch + web search tool (similar to what Claude Code has). Everything I tried had issues: some returned raw HTML, others produced messy markdown full of irrelevant content, and many just broke on JavaScript-heavy sites. Cleaning and post-processing all of that logic became its own problem.
# What problem was I trying to solve? I wanted web tools that return exactly what an LLM should see - clean, relevant, structured data - without developers having to write custom scraping, parsing, or filtering logic for every site.
# How did the approach evolve? This originally started as a small part of another project. While working on it, I realized this piece was actually more feasible and clearly scoped than my original idea, so I decided to pivot and focus on it as a standalone product. The trade-off I chose was simplicity and LLM-ready output over raw speed: the data goes through an LLM layer before being returned, so it’s immediately usable in agents and workflows. It also works across JS-heavy sites and integrates easily with different frameworks.
# Why I’m launching: I’m mainly looking for feedback and testers. I want to know: - Does this solve a real problem for you? - Would you use it in production? - What’s missing?
---
It is free to try out! If you try it out and share feedback, I’m happy to give free credits - just reach out. Any thoughts (good or bad) are hugely valuable at this stage.
👋 Hey Product Hunt! I’m the maker behind Quercle.
# What inspired me to build this?
I was building LLM agents and needed a simple web fetch + web search tool (similar to what Claude Code has). Everything I tried had issues: some returned raw HTML, others produced messy markdown full of irrelevant content, and many just broke on JavaScript-heavy sites. Cleaning and post-processing all of that logic became its own problem.
# What problem was I trying to solve?
I wanted web tools that return exactly what an LLM should see - clean, relevant, structured data - without developers having to write custom scraping, parsing, or filtering logic for every site.
# How did the approach evolve?
This originally started as a small part of another project. While working on it, I realized this piece was actually more feasible and clearly scoped than my original idea, so I decided to pivot and focus on it as a standalone product. The trade-off I chose was simplicity and LLM-ready output over raw speed: the data goes through an LLM layer before being returned, so it’s immediately usable in agents and workflows. It also works across JS-heavy sites and integrates easily with different frameworks.
# Why I’m launching:
I’m mainly looking for feedback and testers. I want to know:
- Does this solve a real problem for you?
- Would you use it in production?
- What’s missing?
---
It is free to try out! If you try it out and share feedback, I’m happy to give free credits - just reach out. Any thoughts (good or bad) are hugely valuable at this stage.
Thanks for checking it out 🙏