Agent MCP

// Prompt Book - Centralized collection of Agent-MCP prompts and workflows export interface PromptTemplate { id: string title: string description: string category: string template: string variables: Array<{ name: string description: string placeholder: string required: boolean }> usage: string examples?: string[] tags: string[] } export interface PromptCategory { id: string name: string description: string icon: string } export const promptCategories: PromptCategory[] = [ { id: 'initialization', name: 'Agent Initialization', description: 'Prompts for setting up admin and worker agents', icon: 'UserPlus' }, { id: 'task-management', name: 'Task Management', description: 'Prompts for creating, assigning, and managing tasks', icon: 'CheckSquare' }, { id: 'context-management', name: 'Context Management', description: 'Prompts for managing project context and memory', icon: 'Database' }, { id: 'debugging', name: 'Debugging & Analysis', description: 'Prompts for troubleshooting and system analysis', icon: 'Bug' }, { id: 'coordination', name: 'Agent Coordination', description: 'Prompts for multi-agent coordination and communication', icon: 'Users' } ] export const promptTemplates: PromptTemplate[] = [ // INITIALIZATION PROMPTS { id: 'admin-init', title: 'Admin Agent Initialization', description: 'Initialize an admin agent with authentication token and context setup', category: 'initialization', template: `Initialize as an admin agent with this token: {{ADMIN_TOKEN}} Please add the MCD.md file to the project context. Don't summarize it.`, variables: [ { name: 'ADMIN_TOKEN', description: 'The admin authentication token from the MCP database', placeholder: 'admin_token_here', required: true } ], usage: 'Use this prompt to set up the main admin agent that will coordinate other agents and manage the project.', examples: [ 'Initialize as an admin agent with this token: abc123def456\nPlease add the MCD.md file to the project context. Don\'t summarize it.' ], tags: ['admin', 'initialization', 'setup', 'auth'] }, { id: 'worker-init', title: 'Worker Agent Initialization', description: 'Initialize a worker agent with specific role and autonomous behavior', category: 'initialization', template: `You are {{AGENT_ID}} agent, your Worker Token: "{{WORKER_TOKEN}}" Look at your tasks and ask the project RAG agent at least 5-7 questions to understand what you need to do. I want you to critically think when asking a question, then criticize yourself before asking that question. How you criticize yourself is by proposing an idea, criticizing it, and based on that criticism you pull through with that idea. AUTO --worker --memory`, variables: [ { name: 'AGENT_ID', description: 'Unique identifier for the worker agent (e.g., frontend-worker, backend-worker)', placeholder: 'frontend-worker', required: true }, { name: 'WORKER_TOKEN', description: 'The specific worker token provided by the admin agent', placeholder: 'worker_token_here', required: true } ], usage: 'Use this prompt in a new AI assistant window to initialize a worker agent that will autonomously work on assigned tasks.', examples: [ 'You are frontend-worker agent, your Worker Token: "xyz789abc123"\n\nLook at your tasks and ask the project RAG agent at least 5-7 questions...' ], tags: ['worker', 'initialization', 'autonomous', 'tasks'] }, { id: 'worker-init-legacy', title: 'Worker Agent Initialization (Legacy)', description: 'Legacy worker initialization format using admin token', category: 'initialization', template: `You are {{AGENT_ID}} agent, your Admin Token: "{{ADMIN_TOKEN}}" Look at your tasks and ask the project RAG agent at least 5-7 questions to understand what you need to do. I want you to critically think when asking a question, then criticize yourself before asking that question. How you criticize yourself is by proposing an idea, criticizing it, and based on that criticism you pull through with that idea. AUTO --worker --memory`, variables: [ { name: 'AGENT_ID', description: 'Unique identifier for the worker agent', placeholder: 'worker-id', required: true }, { name: 'ADMIN_TOKEN', description: 'The admin token (legacy format)', placeholder: 'admin-token', required: true } ], usage: 'Legacy format for worker initialization. Use the newer worker-init template when possible.', tags: ['worker', 'initialization', 'legacy', 'deprecated'] }, // TASK MANAGEMENT PROMPTS { id: 'create-worker', title: 'Create Worker Agent', description: 'Request admin agent to create a new worker agent for specific tasks', category: 'task-management', template: `Create a worker agent with ID "{{AGENT_ID}}" to {{TASK_DESCRIPTION}}.`, variables: [ { name: 'AGENT_ID', description: 'Unique identifier for the new worker agent', placeholder: 'frontend-worker', required: true }, { name: 'TASK_DESCRIPTION', description: 'Brief description of what the worker should accomplish', placeholder: 'implement the login page', required: true } ], usage: 'Tell this to your admin agent to create a new specialized worker for specific tasks.', examples: [ 'Create a worker agent with ID "frontend-worker" to implement the login page.', 'Create a worker agent with ID "api-worker" to build the REST API endpoints.', 'Create a worker agent with ID "database-worker" to design and implement the database schema.' ], tags: ['admin', 'worker-creation', 'task-assignment'] }, { id: 'assign-task', title: 'Assign Task to Agent', description: 'Assign a specific task to an existing agent', category: 'task-management', template: `Assign the following task to {{AGENT_ID}}: Title: {{TASK_TITLE}} Description: {{TASK_DESCRIPTION}} Priority: {{PRIORITY}} {{ADDITIONAL_CONTEXT}}`, variables: [ { name: 'AGENT_ID', description: 'ID of the agent to assign the task to', placeholder: 'frontend-worker', required: true }, { name: 'TASK_TITLE', description: 'Brief title for the task', placeholder: 'Implement user authentication', required: true }, { name: 'TASK_DESCRIPTION', description: 'Detailed description of what needs to be done', placeholder: 'Create login form with validation and JWT token handling', required: true }, { name: 'PRIORITY', description: 'Task priority level', placeholder: 'high', required: false }, { name: 'ADDITIONAL_CONTEXT', description: 'Any additional context or requirements', placeholder: 'Use React hooks and follow existing component patterns', required: false } ], usage: 'Use this to assign specific tasks to agents through the admin agent.', tags: ['task-assignment', 'admin', 'project-management'] }, // CONTEXT MANAGEMENT PROMPTS { id: 'add-context', title: 'Add Project Context', description: 'Add important information to the project context database', category: 'context-management', template: `Please add the {{FILE_NAME}} file to the project context. {{INSTRUCTION}}`, variables: [ { name: 'FILE_NAME', description: 'Name of the file to add to context', placeholder: 'MCD.md', required: true }, { name: 'INSTRUCTION', description: 'Special instructions for how to process the file', placeholder: "Don't summarize it.", required: false } ], usage: 'Use this to add important project files to the shared context database.', examples: [ 'Please add the MCD.md file to the project context. Don\'t summarize it.', 'Please add the API-specification.md file to the project context.', 'Please add the database-schema.sql file to the project context. Include all table structures.' ], tags: ['context', 'documentation', 'knowledge-management'] }, { id: 'rag-query', title: 'Query Project RAG', description: 'Ask questions about the project using the RAG system', category: 'context-management', template: `Ask the project RAG: "{{QUESTION}}"`, variables: [ { name: 'QUESTION', description: 'Question to ask the RAG system about the project', placeholder: 'What is the database schema for user authentication?', required: true } ], usage: 'Use this format to query the RAG system for project-specific information.', examples: [ 'Ask the project RAG: "What is the authentication flow for the application?"', 'Ask the project RAG: "What components have already been implemented?"', 'Ask the project RAG: "What are the API endpoints for user management?"' ], tags: ['rag', 'query', 'knowledge-retrieval'] }, // DEBUGGING PROMPTS { id: 'debug-agent-status', title: 'Debug Agent Status', description: 'Check the status and health of agents in the system', category: 'debugging', template: `Please check the status of all agents and provide a summary of: 1. Active agents and their current tasks 2. Any failed or stuck agents 3. Task completion status 4. System health overview {{SPECIFIC_CONCERNS}}`, variables: [ { name: 'SPECIFIC_CONCERNS', description: 'Any specific issues or concerns to investigate', placeholder: 'Focus on why the frontend-worker seems to be stuck', required: false } ], usage: 'Use this with the admin agent to get a comprehensive system status report.', tags: ['debugging', 'status', 'health-check', 'admin'] }, { id: 'debug-task-flow', title: 'Debug Task Flow', description: 'Analyze task dependencies and execution flow', category: 'debugging', template: `Analyze the task flow for {{TASK_OR_FEATURE}} and identify: 1. Task dependencies and order 2. Any blocking issues 3. Agent assignments and capabilities 4. Estimated completion timeline Please provide recommendations for optimization.`, variables: [ { name: 'TASK_OR_FEATURE', description: 'Specific task or feature to analyze', placeholder: 'user authentication system', required: true } ], usage: 'Use this to analyze complex task dependencies and identify bottlenecks.', tags: ['debugging', 'task-analysis', 'workflow', 'optimization'] }, // COORDINATION PROMPTS { id: 'coordinate-agents', title: 'Coordinate Multiple Agents', description: 'Coordinate work between multiple agents on related tasks', category: 'coordination', template: `Coordinate the following agents to work on {{PROJECT_PHASE}}: Agents involved: {{AGENT_LIST}} Requirements: - {{REQUIREMENT_1}} - {{REQUIREMENT_2}} - {{REQUIREMENT_3}} Please ensure proper communication and dependency management between agents.`, variables: [ { name: 'PROJECT_PHASE', description: 'The project phase or feature being worked on', placeholder: 'user authentication implementation', required: true }, { name: 'AGENT_LIST', description: 'List of agents and their roles', placeholder: '- frontend-worker: UI components\n- backend-worker: API endpoints\n- database-worker: schema design', required: true }, { name: 'REQUIREMENT_1', description: 'First coordination requirement', placeholder: 'Frontend and backend must agree on API contract', required: false }, { name: 'REQUIREMENT_2', description: 'Second coordination requirement', placeholder: 'Database schema must be finalized before API implementation', required: false }, { name: 'REQUIREMENT_3', description: 'Third coordination requirement', placeholder: 'All components must follow the established design system', required: false } ], usage: 'Use this with the admin agent to coordinate complex multi-agent workflows.', tags: ['coordination', 'multi-agent', 'project-management', 'admin'] }, { id: 'handoff-task', title: 'Task Handoff Between Agents', description: 'Transfer a task from one agent to another with context', category: 'coordination', template: `Transfer task "{{TASK_TITLE}}" from {{FROM_AGENT}} to {{TO_AGENT}}. Current Status: {{CURRENT_STATUS}} Handoff Notes: {{HANDOFF_NOTES}} Next Steps: {{NEXT_STEPS}}`, variables: [ { name: 'TASK_TITLE', description: 'Title of the task being transferred', placeholder: 'User Profile Component', required: true }, { name: 'FROM_AGENT', description: 'Agent currently working on the task', placeholder: 'frontend-worker-1', required: true }, { name: 'TO_AGENT', description: 'Agent who will take over the task', placeholder: 'frontend-worker-2', required: true }, { name: 'CURRENT_STATUS', description: 'Current state of the task', placeholder: 'Basic component structure completed, needs styling and validation', required: true }, { name: 'HANDOFF_NOTES', description: 'Important context for the new agent', placeholder: 'Component uses React Hook Form for validation, follow existing pattern in LoginForm', required: true }, { name: 'NEXT_STEPS', description: 'What the new agent should focus on', placeholder: '1. Add form validation\n2. Implement responsive design\n3. Add unit tests', required: true } ], usage: 'Use this to properly transfer tasks between agents with full context preservation.', tags: ['handoff', 'task-transfer', 'coordination', 'context-preservation'] } ] // Helper functions for working with prompts export function getPromptsByCategory(categoryId: string): PromptTemplate[] { return promptTemplates.filter(prompt => prompt.category === categoryId) } export function searchPrompts(query: string): PromptTemplate[] { const lowercaseQuery = query.toLowerCase() return promptTemplates.filter(prompt => prompt.title.toLowerCase().includes(lowercaseQuery) || prompt.description.toLowerCase().includes(lowercaseQuery) || prompt.tags.some(tag => tag.toLowerCase().includes(lowercaseQuery)) || prompt.template.toLowerCase().includes(lowercaseQuery) ) } export function getPromptById(id: string): PromptTemplate | undefined { return promptTemplates.find(prompt => prompt.id === id) } export function fillPromptTemplate(template: string, variables: Record<string, string>): string { let filled = template Object.entries(variables).forEach(([key, value]) => { const regex = new RegExp(`{{${key}}}`, 'g') filled = filled.replace(regex, value) }) return filled } export function getRequiredVariables(prompt: PromptTemplate): string[] { return prompt.variables.filter(v => v.required).map(v => v.name) } export function validatePromptVariables(prompt: PromptTemplate, variables: Record<string, string>): string[] { const errors: string[] = [] const required = getRequiredVariables(prompt) required.forEach(varName => { if (!variables[varName] || variables[varName].trim() === '') { errors.push(`Required variable "${varName}" is missing or empty`) } }) return errors }