Skip to main content
Glama

mcp_instruct_onboarding

Set up your personal profile and AI agent to enable long-term memory and context for AI assistants across sessions.

Instructions

Start the MCP Instruct onboarding process - sets up personal profile and AI agent

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
actionNoOnboarding action to performstart
dataNoOptional data for quick setup

Implementation Reference

  • Main handler for the mcp_instruct_onboarding MCP tool. Handles 'start', 'check_status', and 'quick_setup' actions. Initializes AgentManager, checks if knowledge base is new, provides onboarding messages or status, and performs quick setup by updating personal/professional info and activating agents based on role.
    case 'mcp_instruct_onboarding': {
      await am.initialize();
      const { action = 'start', data } = args as any;
      
      switch (action) {
        case 'start':
          const isNew = km.isNew();
          if (isNew) {
            return {
              content: [
                {
                  type: 'text',
                  text: `🎉 Welcome to MCP Instruct!\n\nI'll help you set up your personal knowledge base and AI agent.\n\nTell me:\n1. Your name\n2. Your role/occupation\n3. Preferred agent mode (IT, Security, Sales, etc.)\n\nOr use quick setup: mcp_instruct_onboarding with action="quick_setup"`
                }
              ]
            };
          } else {
            const kb = km.getKnowledgeBase();
            const activeAgent = am.getActiveAgent();
            return {
              content: [
                {
                  type: 'text',
                  text: `Welcome back, ${kb.personal.name || 'User'}!\n\nProfile: ${kb.professional.occupation || 'Not set'}\nActive Agent: ${activeAgent?.name || 'None'}\n\nYou can switch agents or update your profile anytime.`
                }
              ]
            };
          }
          
        case 'check_status':
          const kb = km.getKnowledgeBase();
          const agent = am.getActiveAgent();
          return {
            content: [
              {
                type: 'text',
                text: JSON.stringify({
                  isNew: km.isNew(),
                  profile: {
                    name: kb.personal.name,
                    role: kb.professional.occupation
                  },
                  activeAgent: agent?.name
                }, null, 2)
              }
            ]
          };
          
        case 'quick_setup':
          if (data?.name) {
            await km.updatePersonal({ name: data.name });
          }
          if (data?.role) {
            await km.updateProfessional({ occupation: data.role });
            
            // Auto-select agent based on role
            if (data.role.toLowerCase().includes('it')) {
              am.setActiveAgent('it-expert');
            } else if (data.role.toLowerCase().includes('security')) {
              am.setActiveAgent('ethical-hacker');
            } else if (data.role.toLowerCase().includes('sales')) {
              am.setActiveAgent('sales-expert');
            }
          }
          if (data?.preferredAgent) {
            am.setActiveAgent(data.preferredAgent);
          }
          
          return {
            content: [
              {
                type: 'text',
                text: `✅ Setup complete!\n\nProfile: ${data?.name} - ${data?.role}\nActive Agent: ${am.getActiveAgent()?.name || 'None'}\n\nYour profile is saved and will persist across sessions.`
              }
            ]
          };
          
        default:
          return {
            content: [
              {
                type: 'text',
                text: 'Unknown onboarding action'
              }
            ]
          };
      }
    }
  • Input schema definition for the mcp_instruct_onboarding tool, specifying parameters for action (start/check/quick_setup) and optional data (name, role, preferredAgent). This schema is used for tool listing and validation in the MCP server.
    {
      name: 'mcp_instruct_onboarding',
      description: 'Start the MCP Instruct onboarding process - sets up personal profile and AI agent',
      inputSchema: {
        type: 'object',
        properties: {
          action: {
            type: 'string',
            enum: ['start', 'check_status', 'quick_setup'],
            default: 'start',
            description: 'Onboarding action to perform'
          },
          data: {
            type: 'object',
            description: 'Optional data for quick setup',
            properties: {
              name: { type: 'string' },
              role: { type: 'string' },
              preferredAgent: { type: 'string' }
            }
          }
        }
      }
    }
  • src/index.ts:423-425 (registration)
    Registration of all tools including mcp_instruct_onboarding via the ListToolsRequestHandler, which returns the complete tools array containing the tool definition.
    server.setRequestHandler(ListToolsRequestSchema, async () => {
      return { tools };
    });

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/hlsitechio/mcp-instruct'

If you have feedback or need assistance with the MCP directory API, please join our Discord server