HACKER Q&A
📣 thienz

When should you stop building an open-source AI agent framework?


Hi HN,

I've been building an open-source Python framework for production-ready AI agents (lightweight, with built-in circuit breakers, multi-LLM support including Ollama, ReAct/ReWOO/ToT reasoning, and safety features like guardrails/idempotency).

Repo: https://github.com/thienzz/Kite (just released v0.1.0 on PyPI: pip install kite-agent)

My questions: - How do you know when to keep going vs. pivot/stop on an open-source project with low initial interest? - What were your experiences with early traction on AI/tools projects? Did any of yours start slow but pick up later? - For those building agents in production: What pain points do you still have that frameworks aren't solving well (reliability, local runs, cost control)?

Would love honest advice – feeling a bit demotivated but still believe in the idea.

Thanks!

  👤 verdverm Accepted Answer ✓
This is an insanely crowded space and the answer depends on your perspective

Are you trying to turn the open source into a money making thing? What makes yours better than the 1000s of other options? Marketing, branding, and luck are probably what determines the winners

Are you making it for yourself and share it just to share it? Don't worry about metrics and build something you love to use yourself. Write docs and make a discord. Projects with 10s of users can be just fine. Getting even one power user in your discord will help the motivation.

So what are your goals?