agentvoy

The universal AI agent platform. Scaffold, develop, and deploy production-ready AI agents in seconds — any framework, any model, secure by default.

npx agentvoy create my-agent

Quick start

# Interactive — choose framework, model, and build mode
npx agentvoy create my-project

# Agent mode with defaults (OpenAI + GPT-4o)
npx agentvoy create my-project --yes

# App mode — API + chat UI + DevTools + Docker
npx agentvoy create my-project --build-mode app --deploy-target docker --yes

# Start development server with DevTools
cd my-project-app
agentvoy dev    # http://localhost:8080/dev

# Deploy to Docker or Fly.io
agentvoy deploy --target docker
agentvoy deploy --target fly-io

Two paths

Agent — fast local development with a flat project structure and interactive REPL.

App — deployable agentic app with FastAPI server, Streamlit chat UI, real-time DevTools dashboard, and cloud deployment configs.

Supported frameworks

| Framework | Language | Status | |-----------|----------|--------| | OpenAI Agents SDK | Python | Available | | Google ADK | Python | Available | | CrewAI | Python | Available | | LangGraph | Python | Available | | Anthropic SDK | Python | Available | | LlamaIndex | Python | Available | | AutoGen | Python | Available |

Supported model providers

OpenAI · Anthropic · Google · Ollama (local) · Groq · Mistral

What gets generated

Agent mode

my-project-agent/
  agent.py          # Agent logic with agentic loop
  tools.py          # Custom tools
  run.py            # Interactive REPL entry point
  agent.guard.yml   # Guardrails & permissions config
  requirements.txt
  .env.example

App mode

my-project-app/
  src/
    agents/agent.py     # Agent logic
    tools/tools.py      # Custom tools
    trace/tracer.py     # Execution tracing
  server.py             # FastAPI — /run, /health, /dev, /ws/trace
  streamlit_app.py      # Chat UI with model picker & glass theme
  devtools.html         # Real-time DevTools dashboard
  Dockerfile
  agent.guard.yml
  requirements.txt

DevTools — agentvoy dev

Start your agent with the live DevTools dashboard:

agentvoy dev

• Real-time trace streaming via WebSocket
• Event timeline: agent_start, llm_call, tool_call, guard_check, pipeline_stage
• Pipeline visualization for multi-agent apps
• Dark-themed single-page dashboard — no extra dependencies

Deploy — agentvoy deploy

One-command deployment to Docker or Fly.io:

agentvoy deploy --target docker    # Build + run locally
agentvoy deploy --target fly-io    # Deploy to cloud
agentvoy deploy --dry-run          # Generate files only

Deployment targets: Docker, Fly.io, Railway, GCP Cloud Run, AWS Lambda.

agent.guard.yml — built-in guardrails

Every project ships with a universal guardrails config:

guardrails:
  input:
    block_prompt_injection: true
    pii_detection: warn
  behavior:
    max_iterations: 20
    cost_limit: "$1.00"
    timeout: 5m
permissions:
  execution:
    allow_shell: false

Enforced at runtime by agentvoy-guard — automatically included in every project.

All commands

agentvoy create [name]     # Create a new agent or app project
agentvoy dev               # Start agent server with DevTools dashboard
agentvoy deploy            # Deploy to Docker, Fly.io, or other targets
agentvoy init              # Add agent.guard.yml to an existing project
agentvoy validate          # Validate your agent.guard.yml
agentvoy list              # List all supported frameworks and providers

Links

• Website
• GitHub
• agentvoy-guard on PyPI

License

Apache 2.0