Skip to content

Summary

Gives coding agents your CI test results from Currents: failed runs, flaky tests, error trends, quarantine rules and Playwright trace evidence.

Features

  • Query CI runs, spec files and individual test results
  • Fetch failure context for agents to fix broken tests
  • Manage quarantine, skip and tag rules and webhooks
  • Collect screenshots, videos and Playwright traces as shareable evidence
  • Create and link Jira issues from failed tests
  • Ships fix-failing-tests and collect-evidence agent skills

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the environment variables with your own values.
  3. Restart the agent — the "Currents MCP Server" tools will be available.
{
  "mcpServers": {
    "currents-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@currents/mcp"
      ],
      "env": {
        "CURRENTS_API_KEY": "your-api-key"
      }
    }
  }
}

Description

The Currents MCP Server connects Cursor, Claude and other MCP clients to the Currents test-reporting dashboard so an agent can work from what actually happened in CI instead of guessing. It exposes dozens of tools for browsing projects, runs, spec files and individual test results, pulling aggregated error and performance metrics, and managing actions such as quarantine, skip and tag rules, plus webhooks. A get-context tool returns the errors, steps and files of a failed run, spec or test, which is the starting point for fixing broken tests. Evidence tools collect screenshots, videos, traces and attachments with signed download URLs and can build shareable links backed by Playwright traces, with a text digest, filmstrip and DOM snapshots. Jira tools create or link issues straight from a failed test. The server also publishes three agent skills as MCP resources: fix-failing-tests, collect-evidence and browser-evidence. It runs through npx with a Currents API key, is licensed Apache-2.0, and suits teams that already report Playwright, Cypress or similar suites to Currents and want agents to triage flaky tests and CI failures.

Related MCP Servers

MCP: Chrome DevTools MCP

by Google Chrome DevTools team

Google's official MCP server for driving and inspecting a real Chrome browser — automation, debugging, and performance traces.

Checkly's hosted MCP server gives an agent read access to your synthetic monitoring account — check health, failing results, test sessions, Rocky AI root-cause analyses, incidents and status pages — over streamable HTTP with OAuth.

A native x64dbg plugin exposing 84 debugging tools and 22 event callbacks over HTTP, so an AI assistant can drive breakpoints, memory and disassembly directly.

MCP: Snag

by Snag

Snag's official remote MCP server — hands a coding agent the session replay, console logs, network trace and annotated screenshots captured for a bug, then takes the fix PR link back.

Browse all MCP servers →