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

Agent language, runtime, and CLI.

Toolang describes agents in .too files, executes them with a runtime, and manages them with a single CLI. It is designed for human-agent collaboration, composable caps, and long-lived agent homes.

Get started

Read the design
example.too
use skill briceyan/pdf-processing
use prompt briceyan/summarize
 
struct WorkspaceSummary: ```json
{
  "summary": "string",
  "highlights": ["string"]
}
```
 
thunk summarize-workspace(user) => WorkspaceSummary:
    tools - memory.forget
    model = gpt-5.3, default
 
    Read the available context.
    Summarize the user's request and the current workspace state.
hello
toolang check summarize-workspace.too
toolang run summarize-workspace
toolang serve summarize-workspace
toolang start summarize-workspace
toolang ps
Language

Learn how agents, caps, homes, and references are described in Toolang.

Runtime

Understand homes, local caps, global storage, hotels, and process tracking.

CLI

Explore commands for checking, running, serving, and managing caps.

Implementation

Review the Python-first stack, parser strategy, model layer, and storage decisions.