Connect the Cloudflare MCP server to Unblocked to give AI agents live Workers context

Debugging a Cloudflare Worker usually means leaving the editor for the Cloudflare dashboard to read invocation logs, then working out which deploy changed what. Cloudflare’s MCP server exposes those Workers logs and metrics to AI agents, and Unblocked connects to it once so every agent can ask about a Worker without its own Cloudflare setup.

This is a third-party MCP server: Cloudflare hosts and maintains it, and Unblocked connects to it as a verified MCP server. Nothing is ingested or indexed — Unblocked calls the read-only tools you select at the moment a question needs them, so answers reflect your Workers as they are now.

Wiring Cloudflare’s MCP server straight into an agent gives it raw log queries. Unblocked’s Context Engine reasons across those results and your other sources — following an error spike to the deploy, the pull request, and the Slack thread behind it — and returns one cited answer.

Once connected, any MCP-compatible coding agent — Claude Code, Cursor, Devin, GitHub Copilot, Codex — reaches Cloudflare context through a single Unblocked MCP server connection.

What you get

• Live Workers logs and metrics — Agents query invocation logs, errors, and metrics as they are right now, through Cloudflare’s own MCP server rather than an index built hours ago.
• One connection for every agent — Connect Cloudflare to Unblocked once. Claude Code, Cursor, Codex, and GitHub Copilot reach it through the Unblocked MCP server, with no per-agent Cloudflare configuration.
• Read-only, tool-level control — Only read-only tools are offered, and you choose which of them Unblocked may call. Change the selection anytime from Settings → Data Sources.
• Context Engine reasoning — Cloudflare results are reconciled with your code, pull requests, incidents, and conversations, so an error in a Worker leads to the change behind it — with citations.

Prerequisites

• An Unblocked account (free to start).
• A Cloudflare account with Workers Logs enabled on the Workers you want to query, and permission to authorize an OAuth app for that account.
• An MCP-compatible coding agent (Claude Code, Cursor, Devin, etc.) to use the context in your editor.

Step 1: Connect Cloudflare to Unblocked

1. In Unblocked, go to Settings → Data Sources → Connect another data source, and under Cloud Infrastructure select Cloudflare.
2. The Cloudflare MCP server URL is filled in for you. Choose OAuth, click Connect, and sign in to Cloudflare to authorize Unblocked.
3. Once Unblocked discovers the server’s tools, select the read-only tools it may use and click Save Settings.

Step 2: Install the Unblocked MCP server

On macOS or Linux, one command auto-detects and configures the supported agents on your machine:

On Windows, or when you prefer remote MCP, 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

Cloudflare hosts and maintains this MCP server. Unblocked calls only the read-only tools you select, at the moment a question needs them:

• query_worker_observabilityQueries Workers Observability to list log events, calculate metrics, and find specific invocations across your Workers.
• observability_keysLists the fields available in your Workers logs — metadata, Worker-specific, and custom logged fields — so queries filter on real field names.
• observability_valuesFinds the values a field takes, so a query can filter on the exact Worker, status, or custom value it needs.

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

Unblocked acts as the MCP context layer for your Cloudflare data, searching Workers logs, invocations, and metrics alongside your other connected sources and returning cited results through these tools:

• context_researchResearches a question across your connected sources and calls the Cloudflare tools you selected when Workers logs or metrics are relevant, returning a synthesized answer with citations.
• context_get_urlsFetches the full content behind URLs your agent encounters across connected sources — Jira, Linear, Notion, Confluence, Slack, GitHub, and more.

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

• Why did errors on the api-gateway Worker spike yesterday, and which pull request shipped just before?
• Which Worker used the most CPU time this week, and who last changed it?
• Compare error rates between the production and staging Workers since the last deploy.
• What do the logs for this failed invocation say, and is there a Slack thread or ticket about it?