VibeCody Feature Demos

Interactive walkthroughs covering VibeCody’s major capabilities across CLI, Desktop IDE, and API surfaces.

Getting Started

Demo Surface Description
01 — First Run & Setup CLI Install, configure providers, first AI chat
02 — TUI Interface CLI Navigate the terminal UI, panels, keybindings

AI & Providers

Demo Surface Description
03 — Multi-Provider AI Chat CLI + UI Switch between 17 AI providers, streaming responses
04 — Agent Loop & Tool Execution CLI + UI Autonomous coding with file edit, shell, search tools
05 — Model Arena UI Side-by-side model comparison and ranking
06 — Cost Observatory CLI + UI Track token usage and costs across providers

Code Intelligence

Demo Surface Description
07 — Inline Chat & Completions UI Context-aware code suggestions in the editor
08 — Code Search & Embeddings CLI + UI Semantic search across codebases
09 — Autofix & Diagnostics CLI + UI Automated bug detection and repair
10 — Code Transforms CLI + UI AST-based refactoring and code generation

DevOps & Infrastructure

Demo Surface Description
11 — Docker & Container Management CLI + UI Build, run, manage containers
12 — Kubernetes Operations CLI + UI Deploy, scale, monitor K8s workloads
13 — CI/CD Pipeline CLI + UI GitHub Actions, pipeline monitoring
14 — Cloud Provider Integration CLI + UI AWS/GCP/Azure scanning, IAM, IaC generation
15 — Deploy & Database CLI + UI Deployment workflows and database management

MCP & Extensions

Demo Surface Description
16 — MCP Server Integration CLI + UI Connect external tool servers
17 — MCP Lazy Loading CLI + UI Scalable tool registry with on-demand loading
18 — MCP Plugin Directory UI Browse, install, rate verified plugins

Collaboration & Context

Demo Surface Description
19 — Context Bundles CLI + UI Shareable context sets for teams
20 — Agent Teams CLI + UI Multi-agent collaboration with roles
21 — CRDT Collaboration UI Real-time multi-user editing
22 — Gateway Messaging CLI AI assistant on 18 platforms (Slack, Discord, etc.)

Testing & Quality

Demo Surface Description
23 — Test Runner & Coverage CLI + UI Run tests, track coverage, generate tests
24 — Red Team Security CLI + UI Security scanning and vulnerability detection
25 — SWE-bench Benchmarking CLI + UI Benchmark AI coding performance
26 — QA Validation Pipeline CLI + UI Multi-round quality validation

Developer Tools

Demo Surface Description
27 — HTTP Playground CLI + UI API testing with history and collections
28 — GraphQL Explorer UI Schema introspection and query building
29 — Regex & Encoding Tools UI Regex tester, JWT decoder, base converter
30 — Notebook & Scripts CLI + UI Interactive notebooks and script runner

Enterprise & Advanced

Demo Surface Description
31 — Batch Builder CLI + UI Generate entire codebases (3M+ lines)
32 — Legacy Migration CLI + UI COBOL/Fortran to modern languages
33 — App Builder CLI + UI Full-stack app scaffolding from prompts
34 — Usage Metering CLI + UI Credit budgets and team cost allocation
35 — Compliance & Audit CLI + UI SOC 2 controls and audit trails

Running Demos

CLI Demos

# Run a specific demo recording
vibecli demo run <demo-id>

# List available demos
vibecli demo list

# Generate a demo with AI
vibecli demo generate --feature "agent loop"

# Export demo as HTML slideshow
vibecli demo export <demo-id> --format html

VibeUI Demos

  1. Open VibeUI: cd vibeui && npm run tauri dev
  2. Navigate to the Demo tab in the AI panel
  3. Browse demos by category
  4. Click Play to step through interactively

Self-Hosted / Air-Gapped

# Run with Ollama (no internet required)
docker-compose up -d
vibecli --provider ollama chat "Hello"

Demo JSON Format

Demos use VibeCody’s DemoRecording format:

{
  "id": "demo-agent-loop",
  "title": "Agent Loop & Tool Execution",
  "description": "Watch the AI agent autonomously edit files",
  "steps": [
    { "action": "Navigate", "target": "http://localhost:7878" },
    { "action": "Type", "target": "#prompt", "value": "Fix the bug in auth.rs" },
    { "action": "Narrate", "value": "The agent analyzes the codebase..." },
    { "action": "Screenshot", "label": "agent-thinking" },
    { "action": "Wait", "duration_ms": 2000 },
    { "action": "Assert", "target": ".tool-call", "value": "contains:EditFile" }
  ],
  "tags": ["agent", "tools", "coding"]
}

Step types: Navigate, Click, Type, Wait, Screenshot, Assert, Narrate, EvalJs, Scroll, WaitForSelector