examples

Bindu Examples

Example agents demonstrating Bindu's capabilities - from simple bots to multi-agent systems with payments.

Quick Start

Prerequisites

• Python 3.12+
• uv package manager
• OpenRouter API key

Setup

git clone https://github.com/getbindu/bindu.git
cd bindu
uv sync --dev --extra agents
export OPENROUTER_API_KEY="your-key-here"  # pragma: allowlist secret

Run an Agent

uv run examples/beginner/echo_simple_agent.py

Agents run on ports 3773-3780 with UI at http://localhost:[port]/docs

You can override the port for any example without editing code:

# Linux/macOS
export BINDU_PORT=4000

# Windows PowerShell
$env:BINDU_PORT="4000"

For full URL override, use BINDU_DEPLOYMENT_URL (e.g. http://127.0.0.1:5001).

Examples

Beginner

• beginner/echo_simple_agent.py - Minimal echo bot
• beginner/beginner_zero_config_agent.py - Zero-config agent with web search
• beginner/agno_simple_example.py - Joke generator
• beginner/agno_example.py - Research assistant with DuckDuckGo
• beginner/faq_agent.py - Documentation search agent
• beginner/agno_notion_agent.py - Notion integration
• beginner/ag2_simple_example.py - AG2 (AutoGen) simple agent
• beginner/minimax_example.py - MiniMax AI research agent (OpenAI-compatible)
• beginner/dspy_agent.py - DSPy framework integration
• beginner/agno_paywall_example.py - Paywall-protected agent
• beginner/echo_agent_behind_paywall.py - Echo agent with payment requirement

Specialized

• summarizer/ - Text summarization agent
• weather-research/ - Weather intelligence agent
• web-scraping-agent/ - AI web scraping agent with ScrapeGraph + Mem0 memory
• premium-advisor/ - Paid agent with X402 payments (0.01 USDC per query)
• news-summarizer/ - Real-time news search and summarization using local Ollama
• document-analyzer/ - PDF/DOCX document analysis and Q&A agent
• speech-to-text/ - Audio transcription using Gemini 2.0 Flash (MP3, WAV, OGG, M4A)
• ai-data-analysis-agent/ - Autonomous data analyst with CSV profiling and visualization
• cybersecurity-newsletter/ - Security news aggregator with CVE tracking

TypeScript (Language-Agnostic via gRPC)

• typescript-openai-agent/ - OpenAI SDK agent bindufied with TypeScript SDK
• typescript-langchain-agent/ - LangChain.js agent bindufied with TypeScript SDK

Runtime Providers

• runtime-boxd-agent/ - Echo agent that runs as a real microservice inside a boxd microVM (own URL, DID, persistent disk). See docs/runtime/ for the runtime-provider abstraction.

TypeScript agents use @bindu/sdk which automatically launches the Bindu Python core in the background. Same A2A protocol, same DID, same everything — just a different language. See the gRPC documentation for details.

Advanced

• agent_swarm/ - Multi-agent collaboration system
• cerina_bindu/cbt/ - CBT therapy protocol generator
• ag2_research_team/ - Multi-agent research pipeline using AG2 (AutoGen)
• langgraph_blog_writing_agent/ - Map-Reduce blog writing with LangGraph
• hermes_agent/ - Nous Research's tool-using coding/research agent (web, file, code-exec) bindufied with tiered safety controls

Components

• skills/ - Reusable agent capabilities

Environment Variables

# Required (at least one LLM provider)
OPENROUTER_API_KEY=sk-or-v1-your-api-key-here
MINIMAX_API_KEY=your-minimax-api-key  # Alternative: MiniMax AI (https://platform.minimaxi.com)

# Optional
PORT=4000
BINDU_PORT=4000
BINDU_DEPLOYMENT_URL=http://localhost:4000
HYDRA__ADMIN_URL=https://hydra-admin.getbindu.com
HYDRA__PUBLIC_URL=https://hydra.getbindu.com
DATABASE_URL=postgresql+asyncpg://user:pass@host/db  # pragma: allowlist secret
REDIS_URL=rediss://default:pass@host:6379  # pragma: allowlist secret

X402 Payments

The premium-advisor/ example shows how to monetize agents with X402 payments:

uv run examples/premium-advisor/premium_advisor.py

Users must pay 0.01 USDC before the agent responds.

Testing

Web UI

cd frontend
npm run dev

API

curl -X POST ${BINDU_DEPLOYMENT_URL:-http://localhost:${BINDU_PORT:-3773}}/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"message/send","params":{...},"id":"1"}'

Building Your Own

from bindu import Agent

agent = Agent(
    name="My Agent",
    description="What it does",
    model="openai/gpt-4o",
)

agent.instructions = ["Behavior guidelines"]

if __name__ == "__main__":
    agent.serve(port=3773)

Documentation

• Bindu Docs
• gRPC Language-Agnostic Guide
• TypeScript SDK
• Payment Guide
• DID Guide
• Skills Guide

Contributing

1. Create your agent in the appropriate folder
2. Add README with usage instructions
3. Include .env.example
4. Submit pull request

License

See LICENSE.md