Method RunAsync

Namespace

LMKit.Agents

Assembly

LM-Kit.NET.dll

RunAsync(string, CancellationToken)

Executes the agent with the given input asynchronously.

public Task<AgentExecutionResult> RunAsync(string input, CancellationToken cancellationToken = default)

Parameters

input string

The user input or task prompt.

cancellationToken CancellationToken

A token to cancel the operation.

Returns

Task<AgentExecutionResult>

A task that resolves to the execution result.

Examples

Running an async task with cancellation:

using LMKit.Model;
using LMKit.Agents;

using var model = new LM("path/to/model.gguf");
var agent = new Agent(model);
agent.Identity = new AgentIdentity("Analyst", "Provide detailed analysis.");

using var cts = new CancellationTokenSource(TimeSpan.FromMinutes(1));

try
{
    var result = await agent.RunAsync(
        "Analyze the pros and cons of remote work.",
        cts.Token
    );
    Console.WriteLine(result.Content);
}
catch (OperationCanceledException)
{
    Console.WriteLine("Task was cancelled.");
}

Remarks

Creates a temporary AgentExecutor for this execution. For multi-turn conversations or repeated executions, create and reuse an executor directly for better performance.

Exceptions

ArgumentNullException

Thrown when input is null or empty.

InvalidOperationException

Thrown when the agent is not properly configured.

RunAsync(string, AgentExecutionOptions, CancellationToken)

Executes the agent with the given input asynchronously using the specified options.

public Task<AgentExecutionResult> RunAsync(string input, AgentExecutionOptions options, CancellationToken cancellationToken = default)

Parameters

input string

The user input or task prompt.

options AgentExecutionOptions

Execution options, or null to use defaults.

cancellationToken CancellationToken

A token to cancel the operation.

Returns

Task<AgentExecutionResult>

A task that resolves to the execution result.

Examples

Running with options and inspecting tool calls:

using LMKit.Model;
using LMKit.Agents;

using var model = new LM("path/to/model.gguf");
var agent = new Agent(model);
agent.EnsureTools().Register(new MySearchTool());
agent.Planning = PlanningStrategy.ReAct;

var options = new AgentExecutionOptions
{
    MaxIterations = 10,
    Timeout = TimeSpan.FromMinutes(2)
};

var result = await agent.RunAsync("Find recent news about AI.", options);

Console.WriteLine($"Response: {result.Content}");
Console.WriteLine($"Tool calls made: {result.ToolCalls.Count}");
Console.WriteLine($"Inference count: {result.InferenceCount}");

foreach (var toolCall in result.ToolCalls)
{
    Console.WriteLine($"  - {toolCall.ToolName}: {toolCall.ResultType}");
}