Welcome to agent-frontier — the home for the best LLM tools, agent frameworks, and production-ready CLIs/services.
We don’t just build experiments. We build the stuff you can ship: clean interfaces, practical defaults, and components that help agents do real work—fast.
- Best-in-class LLM tooling: useful libraries, frameworks, and utilities for building agents.
- Frontier-level agent behavior: tools that move beyond demos and into reliable execution.
- CLI-first mindset: commands you can trust, integrate, script, and run in pipelines.
- Services you can deploy: pragmatic architectures designed for real systems.
- Open by default: public repos, clear docs, and community-driven improvements.
Because this is the place where:
- the highest-signal ideas become usable tools
- tooling quality matters (DX, stability, docs)
- the best CLIs/services are built with the long run in mind
If you’re building with LLMs and need something that just works—this is where you start.
Browse the org repositories to find:
- agent frameworks and orchestration
- prompt/tool runtimes
- developer CLIs for LLM workflows
- APIs/services for agent execution
- evaluation, observability, and reliability utilities
- Star what you want to use
- Watch what you want to track
- Open issues / PRs for improvements, features, and integrations
- Repositories: see the org list
- Issues: open a ticket in the relevant repo
- Contributions: PRs welcome in any repo
agent-frontier: where the frontier becomes usable.