mcp-ai-agent-guidelines
Version:
A comprehensive Model Context Protocol server providing advanced tools, resources, and prompts for implementing AI agent best practices
64 lines • 1.99 kB
TypeScript
/**
* Type definitions for agent orchestration system
*
* @module agents/types
*/
/**
* Definition of an agent in the registry.
* Contains metadata about the agent's capabilities and how to invoke it.
*/
export interface AgentDefinition {
/** Unique name identifier for the agent */
name: string;
/** Human-readable description of the agent's purpose */
description: string;
/** List of capabilities this agent provides */
capabilities: string[];
/** JSON schema for the agent's input */
inputSchema: object;
/** Optional JSON schema for the agent's output */
outputSchema?: object;
/** Name of the MCP tool backing this agent */
toolName: string;
}
/**
* Information about an agent returned in queries.
* Lightweight version of AgentDefinition for listing purposes.
*/
export interface AgentInfo {
/** Unique name identifier for the agent */
name: string;
/** Human-readable description of the agent's purpose */
description: string;
/** List of capabilities this agent provides */
capabilities: string[];
/** Whether the agent is currently available */
available: boolean;
}
/**
* Request to hand off work from one agent to another.
*/
export interface HandoffRequest {
/** Name of the agent initiating the handoff (optional) */
sourceAgent?: string;
/** Name of the target agent to receive the handoff */
targetAgent: string;
/** Context data to pass to the target agent */
context: unknown;
/** Optional reason for the handoff */
reason?: string;
}
/**
* Result of executing a handoff to another agent.
*/
export interface HandoffResult {
/** Whether the handoff was successful */
success: boolean;
/** Output from the target agent */
output: unknown;
/** Time in milliseconds it took to execute the handoff */
executionTime: number;
/** Error message if the handoff failed */
error?: string;
}
//# sourceMappingURL=types.d.ts.map