Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.
Let the model call your functions
RunAnywhere.llm.tools.register(weatherTool, (args) async {
final city = args['city']?.stringValue ?? '';
return {'tempC': ToolValue.number(await fetchTemp(city))};
});
final result = await RunAnywhere.llm.generate('What is the weather in Lisbon?');
void register(ToolDefinition tool, ToolRunner executor)
void unregister(String name)
List<ToolDefinition> list()
clear(). Unregister by name.
final weatherTool = ToolDefinition(
name: 'get_weather',
description: 'Get the current temperature for a city',
parameters: [
ToolParameter(
name: 'city',
type: ParameterType.string,
description: 'City name',
required: true,
),
],
);
| Field | Default | Meaning |
|---|---|---|
tools | empty | Tools for this call only |
toolChoice | auto | Whether the model may call a tool |
maxToolCalls | 5 | Cap on calls in one exchange |
final result = await RunAnywhere.llm.generate(
'Weather in Lisbon and Porto?',
options: LlmOptions(toolChoice: ToolChoice.auto, maxToolCalls: 3),
);
tools.register are offered to every generation. Tools passed in
options.tools apply to that call alone.
ToolRunner is ordinary Dart. It can await, so network and file work are fine, but the
generation is paused while it runs. A slow executor reads to the user as a hung model.
RunAnywhere.llm.tools.unregister('get_weather');
final registered = RunAnywhere.llm.tools.list();