Vibe Guide

Overview

Vibe analyzes natural language prompts and provides appropriate development workflows, checklists, and project guidance. It operates as an MCP server that integrates with AI agents in VS Code.

Installation

Configure Vibe as an MCP server in VS Code by creating .vscode/mcp.json in your workspace:

{
  "servers": {
    "vibe-guide": {
      "type": "stdio",
      "command": "npx",
      "args": ["vibe-guide@latest", "mcp-server"],
      "cwd": "${workspaceFolder}",
      "env": {}
    }
  }
}

Note: After creating this file, restart VS Code to enable the MCP integration.

Usage

Once configured, AI agents can use Vibe to:

• Analyze project requirements and suggest appropriate workflows
• Provide step-by-step guidance for development tasks
• Run quality checklists for code review and validation
• Initialize project configurations
• Monitor workflow progress and session state

Vibe automatically detects project type and adapts its recommendations accordingly.

Features

• MCP server integration for AI agent compatibility
• Pattern-based prompt analysis for workflow recommendations
• Structured workflow guidance with session management
• Nested workflow execution - agents can dynamically add workflows and checklists to running sessions
• Workflow discovery - query available workflows and checklists by pattern or category
• Persistent session state - work continues across agent runs with full context preservation
• Project type detection and adaptation
• Quality checklist validation
• Cross-platform compatibility
• Configurable workflow definitions
• Local execution - runs safely on user's machine, no server required

Nested Workflow Capabilities

Agents can now:

• Query workflows: query_workflows(pattern, category) to discover relevant workflows
• Query checklists: query_checklists(pattern) to find validation checklists
• Add workflows to sessions: add_workflow_to_session(session_id, workflow_name) for dynamic composition
• Add checklists to sessions: add_checklist_to_session(session_id, checklist_name) for nested validation
• Build execution trees: Create nested workflow hierarchies that persist across agent runs
• Reduce token usage: Follow structured plans instead of thinking from scratch

This enables agents to compose complex workflows dynamically while maintaining persistent progress tracking.