Add external websites to Unblocked as context for AI agents

Public documentation can be authoritative context for code that depends on an external API, framework, protocol, or standard. Adding a root URL lets Unblocked research that published material alongside internal sources instead of forcing an agent to guess from local usage alone.

External pages are reconciled with your code, pull requests, docs, and tickets, so an agent can pull the exact section of a vendor doc that explains the behavior it is working with — cited, not guessed.

Once connected, any MCP-compatible coding agent — Claude Code, Cursor, Windsurf, GitHub Copilot, Codex — can pull this context through a single Unblocked MCP server connection.

What you get

• Any public page as context — Add a vendor doc, API reference, standard, or blog by root URL and Unblocked cites it in answers alongside your internal sources.
• Grounded external references — Agents quote the exact page rather than relying on stale training data, so third-party behavior is cited, not guessed.
• Add and remove anytime — Manage the list of URLs from the data sources page whenever your reference material changes.
• Cross-source synthesis — External pages are combined with your code, pull requests, docs, and tickets into a single knowledge graph.

Need screenshots or troubleshooting?This is a concise connection summary. Follow the complete setup guide ↗

Prerequisites

• An Unblocked account (free to start).
• The root URL(s) of the publicly accessible pages you want indexed.
• An MCP-compatible coding agent (Claude Code, Cursor, Windsurf, etc.) to use the context in your editor.

Step 1: Connect External websites to Unblocked

1. In Unblocked, go to Settings → Data Sources → Connect another data source, and under Documentation select Websites.
2. Enter the root URL(s) you want indexed — a docs site, API reference, blog, or any public page.
3. Click Save Settings — Unblocked begins incorporating those pages into its answers.
4. Return to this page anytime to add or remove URLs.

Step 2: Install the Unblocked MCP server

On macOS, one command auto-detects and configures your supported agents (Claude Code, Cursor, Windsurf, GitHub Copilot, Codex, VS Code):

The install script supports macOS only. On Windows or Linux, add the server manually using the configuration below, or follow the manual install guide ↗.

claude mcp add -s user --transport http unblocked https://getunblocked.com/api/mcpsse

{
  "mcpServers": {
    "unblocked": {
      "url": "https://getunblocked.com/api/mcpsse"
    }
  }
}

https://getunblocked.com/api/mcpsse

To give Claude Code, Cursor, GitHub Copilot, Codex, or another MCP-compatible AI coding agent access to External websites context, connect External websites to Unblocked and add the single Unblocked MCP server to your agent.

Unblocked acts as the MCP context layer for your External websites data, searching web pages alongside your other connected sources and returning cited results through these tools:

• context_researchMulti-source research across the external web pages you add, plus your code, pull requests, and other connected sources. Returns a synthesized answer with citations back to the page it used.
• context_get_urlsFetches the full content behind a web page URL your agent encounters — plus GitHub, Jira, Linear, Confluence, Slack, and more.

Questions you (or your agent) can ask Unblocked about your External websites data:

• What does the vendor’s API reference say about rate limits?
• What are the required headers per the payment provider’s docs?
• What does the framework changelog say about the breaking change in v3?
• What does the standard specify for this date format?