Scrape.ly — Claude Code MCP server: features and setup
Scrape.ly is an MCP server for Claude Code that finds Actors in the Scrape.ly Store, runs them, and retrieves their data. It can also search Scrape.ly and Crawlee docs.
MIT updated 2026-11-11
claude mcp add --env SCRAPELY_TOKEN=<SCRAPELY_TOKEN> --env SCRAPELY_API_BASE_URL=<SCRAPELY_API_BASE_URL> --transport stdio scrapely -- uvx scrapely-mcpWhat Scrape.ly does
Scrape.ly lets an agent browse Actors, inspect their README files and input schemas, start runs, and retrieve outputs. Results can come from datasets, key-value stores, and logs. The server also searches Scrape.ly and Crawlee documentation and manages saved tasks and automations.
Connect to the hosted HTTP endpoint or run the stdio server locally. Hosted access uses OAuth or a Scrape.ly token; local setup requires Python 3.11 or newer. Provide the token through SCRAPELY_TOKEN or a Scrape.ly credentials file. Actor search, Actor detail lookup, and documentation tools can work without a token.
What it helps with
- find an Actor and inspect its input schema
- run an Actor and retrieve dataset records
- check a run's status and read its logs
- search Scrape.ly or Crawlee documentation
- manage saved Scrape.ly tasks and automations
Example prompts for the agent
- Find an Actor for product data collection and show its input schema.
- Run the Actor and return records from its dataset.
- Search the Scrape.ly docs for instructions on managing a saved task.
How to install Scrape.ly
1Platform CLI
claude mcp add --env SCRAPELY_TOKEN=<SCRAPELY_TOKEN> --env SCRAPELY_API_BASE_URL=<SCRAPELY_API_BASE_URL> --transport stdio scrapely -- uvx scrapely-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from scrapely-mcp <command>
2Platform CLI: remote server
claude mcp add --transport http scrapely https://mcp.scrape.ly1Platform CLI: remote server
codex mcp add scrapely --url https://mcp.scrape.ly2Platform CLI
codex mcp add scrapely --env SCRAPELY_TOKEN=<SCRAPELY_TOKEN> --env SCRAPELY_API_BASE_URL=<SCRAPELY_API_BASE_URL> -- uvx scrapely-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from scrapely-mcp <command>
Config file
{
"mcpServers": {
"scrapely": {
"url": "https://mcp.scrape.ly"
}
}
}{
"mcpServers": {
"scrapely": {
"command": "uvx",
"args": [
"scrapely-mcp"
],
"env": {
"SCRAPELY_TOKEN": "<SCRAPELY_TOKEN>",
"SCRAPELY_API_BASE_URL": "<SCRAPELY_API_BASE_URL>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from scrapely-mcp <command>
Config file
{
"servers": {
"scrapely": {
"type": "http",
"url": "https://mcp.scrape.ly"
}
}
}{
"servers": {
"scrapely": {
"type": "stdio",
"command": "uvx",
"args": [
"scrapely-mcp"
],
"env": {
"SCRAPELY_TOKEN": "<SCRAPELY_TOKEN>",
"SCRAPELY_API_BASE_URL": "<SCRAPELY_API_BASE_URL>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from scrapely-mcp <command>
1Platform CLI: remote server
gemini mcp add -t http scrapely https://mcp.scrape.ly2Platform CLI
gemini mcp add -e SCRAPELY_TOKEN=<SCRAPELY_TOKEN> -e SCRAPELY_API_BASE_URL=<SCRAPELY_API_BASE_URL> scrapely uvx scrapely-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from scrapely-mcp <command>
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Scrape.ly
How do I install Scrape.ly in Claude Code?
For the hosted server, run claude mcp add --transport http scrapely https://mcp.scrape.ly, then authenticate with claude mcp login scrapely.
What does Scrape.ly do?
It lets an agent find and run Actors, then read outputs from datasets, key-value stores, and logs. It also provides tools for documentation, saved tasks, and automations.
What does Scrape.ly need to run?
The hosted endpoint uses OAuth or a Scrape.ly token. The local stdio server requires Python 3.11 or newer, and its token can come from SCRAPELY_TOKEN or a credentials file.