I built a free and open-source tool to evaluate LLM agents: define your agent in YAML, run it against multiple LLMs in parallel, see exactly where each model fails and what it costs.
Browser-native. No install. No account.
https://t.co/KnBXgmPi32
#llm#aiagents#opensource
New update to https://t.co/N0znI4lNPs:
AI credits are now only deducted after a successful inference.
If an agent chain fails or aborts, you won’t be billed for incomplete results.
#buildinpublic#SaaS
I noticed an unclaimed verification token in the DB.
Reached out to the user. He was unsure about magic links.
Explained how it works, but realized the point.
Learning: listen to hesitation. It shows you where to build trust.
I’ve been dogfooding https://t.co/f3nSMRkX4X from day one. Every feature I ship gets tested on own event data first.
What I’ve learned so far:
- You catch blind spots way faster
- “Next Steps” feel real when you actually use them
- Small wins keep momentum high
#buildinpublic
Cooked today at https://t.co/N0znI4lNPs:
• Integrated the AI tutor with docs as its knowledge base
• Fixed the damn dark mode (finally)
• Polished login & verification UX
• Improved inference cost tracking for users
One step at a time 🛠️
#buildinpublic
After 3 months of setbacks, zero motivation, and thinking about quitting. I kept showing up. Rebranded, redesigned the pricing plans, and deployed to production.
What a feeling.
Building in public also means pushing through when no one’s watching.
#buildinpublic
I’ve been quiet for a while and have been cooking...
Rebranded https://t.co/CXIOgFBD48 → https://t.co/N0znI4lNPs
Integrated payments
Added a Free Plan so you can try it instantly
First version is now live & ready to test
#buildinpublic#indiehacker#saas#startups