๐Ÿงญ Why this stack

An honest take on what's actually broken about long-running AI agents โ€” and why fixing it doesn't need a framework.

The problem isn't agents. It's the boring parts.

Most agent failures aren't reasoning failures. They're plumbing failures:

None of this is glamorous. None of it is novel research. All of it is wrecking your agents in production right now.

Five small fixes

fit ยท guard ยท snap ยท vet ยท cast

What it isn't

This isn't a framework. There's no orchestration layer. No vendor lock-in. Each library is one job, ~200 lines, no runtime dependencies. Use one or use all five. Drop it into LangChain or LlamaIndex or your own loop โ€” doesn't care.

BYO LLM

Nothing in here calls a model. vet doesn't know what an LLM is. fit doesn't either โ€” it just counts tokens. cast assumes you've already gotten a string back from somewhere. You wire it up.

Why MIT, why no sponsorship

MIT, all the way down โ€” Python, JavaScript, MCP servers, datasets. No commercial tier. No paid features. Take it, fork it, embed it in a closed-source product if you want.

Sponsor links are intentionally absent โ€” the author is on H1B, taking outside payments isn't allowed. Best contribution is filing an issue when something breaks, or starring the repos.

Try it