---
title: "OpenAI Agents SDK · Scrape.ly Docs"
description: "Use Scrape.ly tools in agents built with the OpenAI Agents SDK."
url: https://scrape.ly/docs/openai-agents-sdk
markdown: https://scrape.ly/docs/openai-agents-sdk.md
---

# OpenAI Agents SDK

> Use Scrape.ly tools in agents built with the OpenAI Agents SDK.

Give agents built with the OpenAI Agents SDK Scrape.ly tools: search the Store, run Actors, wait for their runs, and read the results. Agents connect to the hosted [MCP server](https://scrape.ly/docs/mcp.md), so there is no Scrape.ly package to install.

```bash
pip install openai-agents
```

Create an API token in the [dashboard](https://my.scrape.ly) under **Settings → API & Integrations** and set it as `SCRAPELY_TOKEN`. The agent can do anything the token can, including starting paid runs, so give it a scoped key: `actors:run`, `runs:read`, and `datasets:read`, with a **Max charge per day**. See [API keys & permissions](https://scrape.ly/docs/api-keys.md).

## Give an agent Scrape.ly tools

Open the server with `MCPServerStreamableHttp` and add it to the agent's `mcp_servers`. The SDK lists the server's tools when the agent runs and calls them as the agent decides.

```python
import asyncio
import os

from agents import Agent, Runner
from agents.mcp import MCPServerStreamableHttp


async def main():
    async with MCPServerStreamableHttp(
        name="Scrape.ly",
        params={
            "url": "https://mcp.scrape.ly",
            "headers": {"Authorization": f"Bearer {os.environ['SCRAPELY_TOKEN']}"},
            "timeout": 30,
        },
        cache_tools_list=True,
    ) as scrapely:
        agent = Agent(
            name="Web researcher",
            instructions="Find the right Actor in the Scrape.ly Store, run it, and report what its results show.",
            mcp_servers=[scrapely],
        )
        result = await Runner.run(
            agent,
            "Find an Actor that scrapes Google Maps, run it for 'coffee shops in Berlin' with at most 20 results, "
            "and list the name, rating, and address of each place.",
        )
        print(result.final_output)


asyncio.run(main())
```

The agent uses the SDK's default model unless you set `model` on it. See [MCP in the Agents SDK](https://openai.github.io/openai-agents-python/mcp/) for more options.

## What the agent can do

The agent gets the MCP server's default tools. The ones used most:

| Tool | What it does |
| --- | --- |
| `search-actors` | Find Actors in the Store by keywords. |
| `fetch-actor-details` | Read an Actor's input schema, README, and pricing before running it. |
| `call-actor` | Run an Actor with a JSON input, optionally with `maxItems` and `maxTotalChargeUsd`. |
| `get-actor-run` | Wait for a run and check its status. |
| `get-actor-output` | Read a run's results, paged and kept compact. |

The full list, including saved inputs and automations, is on the [MCP server](https://scrape.ly/docs/mcp.md#tools) page. To run one Actor you already know, name it in the prompt, for example `Run username/actor-name`, so the agent skips the search.

**Costs**: runs started by the agent are billed like any run. Put the limits in the prompt (such as "at most 20 results"), and keep a **Max charge per day** on the key.

---

- Previous: [Mastra](https://scrape.ly/docs/mastra.md)
- Next: [Google ADK](https://scrape.ly/docs/google-adk.md)
- All docs: [Docs index](https://scrape.ly/docs.md), [llms.txt](https://scrape.ly/llms.txt)