Connect the Statsig MCP server to Unblocked to give AI agents live feature gate and experiment context

A flag check in the code says little on its own. Whether a gate is still rolling out, which experiment it belongs to, and how that experiment is performing all live in the Statsig console, so an agent editing gated code is guessing. Statsig’s MCP server exposes gates, experiments, dynamic configs, metrics, and logs to AI agents, and Unblocked connects to it once so every agent can ask about them without its own Statsig setup.

This is a third-party MCP server: Statsig 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 gates and experiments as they are configured now.

Wiring Statsig’s MCP server straight into an agent gives it raw gate and experiment lookups. Unblocked’s Context Engine reasons across those results and your other sources — tracing an experiment back to the pull request that added its gate check, the Slack thread where the team agreed on the ship criteria, and the ticket tracking cleanup — and returns one cited answer.

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

What you get

• Live gate and experiment state — Agents read the current rules of a feature gate, the setup of an experiment, and its results straight from Statsig, so they know what is actually rolled out before touching gated code.
• Results next to the code that ships them — Experiment results, metric definitions, and Logs Explorer data sit alongside the pull requests that introduced each variant, so a question about impact starts from the change itself.
• Read-only, tool-level control — Only read-only Statsig tools are offered — nothing that creates, edits, or reviews a gate, experiment, or config. You choose which of them Unblocked may call from Settings → Data Sources.
• Context Engine reasoning — Statsig results are reconciled with your code, pull requests, tickets, and conversations, so a stale gate leads to the launch decision behind it — with citations.

Prerequisites

• An Unblocked account (free to start).
• A Statsig project and a Statsig account with access to it, able to authorize the Statsig MCP server through OAuth.
• An MCP-compatible coding agent (Claude Code, Cursor, Devin, etc.) to use the context in your editor.

Step 1: Connect Statsig to Unblocked

1. In Unblocked, go to Settings → Data Sources → Connect another data source, and under Data & Analytics select Statsig.
2. The Statsig MCP server URL is filled in for you. Choose OAuth, click Connect, and sign in to Statsig 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

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

• Get_List_of_GatesReturns the feature gates in your Statsig project.
• Get_Gate_Details_by_IDReturns the configuration and rules of a single feature gate.
• Get_List_of_ExperimentsReturns the experiments in your Statsig project.
• Get_Experiment_Overall_ResultsReturns the overall results of an experiment.
• Query_Logs_ExplorerQueries Logs Explorer data in your Statsig project.
• Get_Audit_LogsReturns audit log entries showing who changed which Statsig resources and when.

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

Unblocked acts as the MCP context layer for your Statsig data, searching feature gates and experiments alongside your other connected sources and returning cited results through these tools:

• context_researchResearches a question across your connected sources and calls the Statsig tools you selected when gates, experiments, or their results 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 Statsig data:

• Is the new_checkout_flow gate fully rolled out, and which pull request added the check for it?
• How is the onboarding experiment performing, and what did the team decide in Slack about its ship criteria?
• Which gates referenced in this service look stale, and is there a ticket to clean them up?
• Who last changed the pricing_page dynamic config, and does that line up with the deploy that broke the banner?