Are you an LLM? Read llms.txt for a summary of the docs, or llms-full.txt for the full context.
Skip to content

toolang

A description language and runtime for agents

Try now
uvx toolang run https://toolang.ai/dev.too

Tool calling turned LLMs from chatbots into agents.
Toolang makes agents easy to build, run, and share.

Run shared agents in seconds

You can run agents from your team or the community. With uvx and a model API key, you can get started in seconds.

Chat with your agents

Toolang prints a WebUI link when an agent starts. Open it in your browser to chat, inspect runs, and manage the agent.

Share agents, not setup guides

A .too file is both a readable description and a runnable definition of an agent. Instead of writing a long setup guide, define the agent once in near-natural language — then run it, share it, and let others build on top.

dev.too
use psyche  briceyan/senior-engineer
use psyche  briceyan/concise
use service briceyan/github
use skill   briceyan/codebase-navigation
 
thunk:
  models = gpt-5
  tools = shell, filesystem, web_search, service_use
 
  user:
    Work as a coding agent for software engineering tasks.
    Understand the repository before editing.
    Make focused changes that fit the existing code.
    Verify the result and report exactly what passed.

Create agents, extend with caps

An agent created from scratch works well out of the box. Extend it with caps — composable agent primitives created by you or shared by others — to fit your needs. Or clone an existing agent to start from a proven setup.

Go beyond chat

You don’t have to start a chat every time you want to get work done. Use chores for recurring work, tasks for team collaboration on projects, and scripts to run agentic procedures inside existing workflows like Makefiles and GitHub Actions.

Curious how these pieces fit together?

© 2026 Toolang