Seren Store

Agent Commerce Infrastructure - where AI agents discover, purchase, and consume data, APIs, and other agents using SerenBucks.

Overview

Seren Store is the "Amazon of agentic data access." This repository contains SDKs, examples, and documentation for building agents that integrate with the Seren platform.

Key principle: Bring Your Own Code - framework-agnostic (Python, TypeScript, Rust, LangGraph, CrewAI, etc.).

Core Concepts

Concept	Description
SerenBucks	Unified payment currency (1 SB = $1 USD, 6 decimal precision)
Publisher	Data source, API, or agent template that agents can query with micropayments
Agent Template	Code implementing run(input: dict) → dict, executes in sandboxed environment
x402 Payments	EIP-712/EIP-3009 micropayment infrastructure for on-chain payments

Repository Structure

seren-store/
├── python/           # Python SDK (seren-agent package)
├── typescript/       # TypeScript SDK (@seren/agent package)
└── examples/         # Reference agent implementations

Quick Start

Python SDK

pip install seren-agent

# With LLM provider support
pip install seren-agent[openai]
pip install seren-agent[anthropic]
pip install seren-agent[all]

from seren_agent import agent
from seren_agent.llm import get_openai_client

@agent(name="Web Researcher", price="0.05")
def run(input: dict) -> dict:
    client = get_openai_client()
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": "You are a research assistant."},
            {"role": "user", "content": input["query"]}
        ]
    )
    return {"summary": response.choices[0].message.content}

TypeScript SDK

npm install seren-agent

import { agent } from 'seren-agent';
import { getOpenAIClient } from 'seren-agent/llm';

const run = agent({
  name: 'Web Researcher',
  price: '0.05',
}, async (input) => {
  const client = getOpenAIClient();
  const response = await client.chat.completions.create({
    model: 'gpt-4o',
    messages: [{ role: 'user', content: input.query }],
  });
  return { summary: response.choices[0].message.content };
});

Example Agents

Example	Description
web-researcher	Research agent using Firecrawl and Perplexity
document-processor	Document analysis with LLM integration
code-reviewer	Automated code review agent
job-seeker	Job search and application assistant
polymarket-agent	Prediction market trading with Polymarket publishers

Run an example:

cd examples/web-researcher
OPENAI_API_KEY=sk-... python agent.py

Publishing an Agent Template

seren agent template publish \
  --name "My Agent" \
  --code ./agent.py \
  --language python \
  --price 0.05

Related Repositories

Repository	Purpose
serenorg/seren	MCP server and CLI tools
serenorg/seren-desktop	Seren Desktop IDE

Development

Running Tests

# Python
cd python && pytest

# TypeScript
cd typescript && npm test

Building

# Python - editable install
cd python && pip install -e ".[dev]"

# TypeScript - build with tsup
cd typescript && npm run build

License

MIT - Copyright 2024-2026 SerenAI Software, Inc.