MrScraper connects Google Antigravity to a hosted MCP server for retrieving public web pages as HTML, Markdown, or screenshots, web search, and account usage. The plugin also includes three focused skills that help the agent choose the right workflow and preserve raw page content.
- The hosted Streamable HTTP endpoint at
https://mcp.mrscraper.com/mcp. - OAuth 2.1 authentication through Antigravity's MCP client; no credential is stored in this repository.
- Three skills:
mrscraper,mrscraper-fetch, andmrscraper-serp. - A retrieve-first workflow that keeps raw page responses available for analysis, verification, and follow-up transformations.
The plugin works in Antigravity 2.0, the Antigravity CLI (agy), and the
standalone Antigravity IDE.
- A current version of Google Antigravity.
- A MrScraper account.
- Permission to access and process the target content.
Clone this repository first:
git clone --depth 1 https://github.com/mrscraper-com/mrscraper-antigravity-plugin.gitagy plugin install ./mrscraper-antigravity-plugin/plugins/mrscraper
agy plugin listThe CLI copies the plugin to ~/.gemini/config/plugins/mrscraper/, the shared
configuration directory that Antigravity 2.0 and the Antigravity IDE also read.
Copy the plugin folder to the global plugin directory to use it in every workspace:
mkdir -p ~/.gemini/config/plugins
cp -R mrscraper-antigravity-plugin/plugins/mrscraper ~/.gemini/config/plugins/To enable it for one project only, copy the folder to
.agents/plugins/mrscraper/ in that workspace instead. Restart Antigravity,
then confirm that MrScraper appears under Customizations.
Antigravity must authorize the hosted MCP server once before the agent can use its tools. MrScraper supports dynamic client registration, so no client ID or secret is required.
- Antigravity 2.0 and Antigravity IDE: open Agent settings with
Cmd+,(macOS) orCtrl+,(Windows and Linux), go to Customizations, and select Authenticate next to the MrScraper server. - Antigravity CLI: run
/mcp, select the MrScraper server, and choose Authenticate.
Approve access on the MrScraper consent page in your browser, copy the authorization code that the browser then shows, paste it back into Antigravity, and submit it. Antigravity stores and refreshes the resulting tokens. Never paste OAuth tokens or API keys into the chat.
Fetch https://www.scrapethissite.com/pages/simple/ and summarize the page.
The connection should expose these MCP tools:
| Tool | Purpose |
|---|---|
scraperAPI |
Retrieve a known page as HTML, Markdown, a screenshot, or a combination. |
webSearch |
Discover public pages through web search. |
status |
Inspect subscription status, token usage, and remaining tokens. |
Antigravity asks for approval before running MCP tools unless your permission
policy allows them. In the CLI, the skills are also available as slash
commands, such as /mrscraper:mrscraper-fetch.
When a public URL is already known, the skills direct the agent to retrieve it
with scraperAPI first and treat the raw response as the source of truth. The
agent can read, summarize, compare, or derive structured output locally without
losing details.
For roughly 100 known pages that share a layout, the agent can retrieve pages concurrently, retain every raw response, and apply one reusable local extractor.
To add only the MCP server, put this entry in ~/.gemini/config/mcp_config.json
(all workspaces) or .agents/mcp_config.json (one workspace). In the
Antigravity IDE, open it from … > MCP Servers > Manage MCP Servers >
View raw config.
{
"mcpServers": {
"mrscraper": {
"serverUrl": "https://mcp.mrscraper.com/mcp"
}
}
}From the CLI, the same entry can be written with:
agy mcp add mrscraper https://mcp.mrscraper.com/mcpAntigravity requires serverUrl for remote servers. Then sign in as described
above.
Use OAuth whenever possible. If Antigravity cannot complete OAuth, create an
API key and send it as a bearer token.
An API key carries full account authority, and Antigravity stores this header
in plain text in mcp_config.json, so keep it in your user-level file, never in
a workspace file that is committed.
With the key exported as MRSCRAPER_API_KEY in your shell:
agy mcp add --header "Authorization: Bearer $MRSCRAPER_API_KEY" \
mrscraper https://mcp.mrscraper.com/mcpOr edit ~/.gemini/config/mcp_config.json directly:
{
"mcpServers": {
"mrscraper": {
"serverUrl": "https://mcp.mrscraper.com/mcp",
"headers": {
"Authorization": "Bearer YOUR_MRSCRAPER_API_KEY"
}
}
}
}If you use the API key fallback, skip the plugin, whose server uses OAuth, and
copy only its skills to ~/.gemini/config/skills/, which Antigravity 2.0, the
IDE, and the CLI all read, or to .agents/skills/ in one workspace:
mkdir -p ~/.gemini/config/skills
cp -R mrscraper-antigravity-plugin/plugins/mrscraper/skills/* ~/.gemini/config/skills/The plugin sends MCP tool inputs, including target URLs and search queries, to MrScraper's hosted service. Page responses and tool results are then available to the agent for the requested task. Use it only with public or otherwise authorized content, and follow the target site's requirements.
Review MrScraper's MCP documentation, Privacy Policy, and Acceptable Use Policy before use.
Pull the latest version and reinstall the plugin:
git -C mrscraper-antigravity-plugin pull
agy plugin install ./mrscraper-antigravity-plugin/plugins/mrscraperDisable or remove it with agy plugin disable mrscraper or
agy plugin uninstall mrscraper, or use the toggle and Uninstall action in
Customizations > Installed.
- Product help: MrScraper MCP documentation
- Bugs and feature requests: GitHub Issues
- Account help: support@mrscraper.com
- Security reports: see SECURITY.md
plugins/mrscraper/
├── plugin.json
├── mcp_config.json
└── skills/
Validate the plugin before a release:
agy plugin validate plugins/mrscraperThe skills are copied from the canonical MrScraper MCP skills. Record user-visible changes in CHANGELOG.md and use PUBLISHING.md for the release and Antigravity Marketplace submission checklist.