Class AgentExecutionOptions
Configuration options for agent execution, providing fine-grained control over context sizing, iteration limits, timeouts, and output constraints.
public sealed class AgentExecutionOptions
- Inheritance
-
AgentExecutionOptions
- Inherited Members
Examples
Using execution options for different scenarios:
using LMKit.Model;
using LMKit.Agents;
using var model = new LM("path/to/model.gguf");
var agent = new Agent(model);
// Quick response with timeout
var quickOptions = new AgentExecutionOptions
{
Timeout = TimeSpan.FromSeconds(10),
MaxCompletionTokens = 200
};
var quickResult = agent.Run("Give a one-sentence summary of AI.", quickOptions);
// Deep research with more iterations
var researchOptions = new AgentExecutionOptions
{
MaxIterations = 20,
Timeout = TimeSpan.FromMinutes(5)
};
var researchResult = agent.Run("Analyze the pros and cons of microservices.", researchOptions);
Remarks
Execution options allow you to customize how an agent processes a specific task without modifying the agent's base configuration. Options can override agent defaults for individual executions.
Common Use Cases
- Set timeouts for user-facing applications.
- Limit token output for summary tasks.
- Increase iterations for complex research tasks.
- Clear history for stateless interactions.
Properties
- ClearHistoryBeforeExecution
Gets or sets a value indicating whether to clear conversation history before execution.
- ContextSize
Gets or sets the context size for the conversation.
- Default
Gets the default execution options.
- MaxCompletionTokens
Gets or sets the maximum number of tokens for completion per turn.
- MaxIterations
Gets or sets the maximum number of iterations allowed during execution.
- Timeout
Gets or sets the maximum time allowed for execution.
Methods
- Clone()
Creates a copy of this options instance.