mirror of
https://github.com/andy-stack/vaultkeeper-ai.git
synced 2026-07-22 16:30:27 +00:00
ab9ee08281
Restructure the AI workflow from a single-agent model to a specialized multi-agent system with distinct roles: - Add AgentType enum (Main, Orchestration, Planning, Execution) to define agent specializations - Replace AIControllerService and AIFunctionService with modular agent classes in Services/AIServices/: - MainAgent: Handles user interaction and delegates to orchestration - OrchestrationAgent: Coordinates plan execution and step-by-step workflow - PlanningAgent: Creates and revises execution plans - ExecutionAgent: Executes individual plan steps - AIController: Base class providing common agent loop functionality - Create specialized prompts (OrchestrationPrompt, ExecutionPrompt) for agent-specific behavior - Add OrchestrationResult type to communicate workflow control decisions (continue, abort, replan) - Introduce agent-specific function scoping: - ExecuteWorkflow for main agent - CompleteTask for execution agent - CompleteStep/Replan/CancelPlan for orchestration agent - SubmitPlan/AskUserQuestionPlanning for planning agent - Update BaseAIClass to use agentType property instead of isPlanningAgent boolean for model selection - Simplify ExecutionPlan and ExecutionStep types by removing execution state tracking (moved to agent coordination) - Remove PlanningEnabledAppendix and ExecutionStatus enum (superseded by agent architecture) - Add comprehensive integration tests for agent workflows This architecture provides better separation of concerns, clearer agent responsibilities, and more robust plan execution with explicit orchestration control flow.
42 lines
No EOL
3.1 KiB
TypeScript
42 lines
No EOL
3.1 KiB
TypeScript
import { AIFunction } from "Enums/AIFunction";
|
|
import type { IAIFunctionDefinition } from "../IAIFunctionDefinition";
|
|
|
|
export const ExecuteWorkflow: IAIFunctionDefinition = {
|
|
name: AIFunction.ExecuteWorkflow,
|
|
description: `Delegates a complex goal to a specialized workflow system that will autonomously plan, execute, and return the completed result.
|
|
|
|
When called, this tool hands off the goal to a planning and execution engine that operates independently. The system will analyze the goal, create a detailed plan, execute each step programmatically, handle any necessary replanning if steps fail, and return the final outcome. You do not need to manage or monitor the execution — the result will be returned to you once the workflow completes or if it fails irrecoverably.
|
|
|
|
This is the appropriate choice when:
|
|
- The user's request involves multiple coordinated operations or phases
|
|
- The task requires vault exploration, analysis, or search before acting
|
|
- The optimal approach is unclear and benefits from systematic planning
|
|
- Work spans multiple vault areas or requires sequential dependencies
|
|
- The task is substantial enough that delegating execution is more efficient than step-by-step guidance
|
|
|
|
Do NOT use this tool for:
|
|
- Simple, single-step operations you can perform directly (e.g., reading one file, creating one note)
|
|
- Tasks where you already know the exact steps and can execute them yourself
|
|
- Requests that require real-time user interaction or clarification during execution
|
|
- Operations where partial results need to be shown to the user incrementally
|
|
|
|
The workflow system specializes in breaking down complex tasks, conducting exploratory searches, managing dependencies, recovering from failures through replanning, and coordinating multi-step operations to completion.`,
|
|
parameters: {
|
|
type: "object",
|
|
properties: {
|
|
goal: {
|
|
type: "string",
|
|
description: "The high-level goal to be planned and executed. Should clearly describe the desired end state or outcome. The workflow system will determine the specific steps needed to achieve this. Examples: 'Create a comprehensive summary of all machine learning notes', 'Reorganize project notes by topic and update all internal links', 'Research and compile information about quantum computing from the vault into a new reference document'"
|
|
},
|
|
context: {
|
|
type: "string",
|
|
description: "Optional additional context to inform planning and execution. Include any constraints, preferences, relevant file paths, or domain knowledge that would help produce a better result. Examples: 'User prefers daily notes in YYYY-MM-DD format', 'Focus only on notes created in the last month', 'Preserve existing wiki-link structure and tag conventions'"
|
|
},
|
|
user_message: {
|
|
type: "string",
|
|
description: "A brief message displayed to the user indicating that a workflow is being executed on their behalf. Keep this concise and informative. Examples: 'Organizing your project notes by topic', 'Compiling your AI research into a summary document', 'Restructuring your vault's folder hierarchy'"
|
|
}
|
|
},
|
|
required: ["goal", "user_message"]
|
|
}
|
|
} |