Coding agents and access

MCP server integration

TesterArmy's hosted MCP server gives Claude, Cursor, Codex and VS Code direct access to projects, tests and runs.

START TESTING FOR FREE READ THE SETUP DOCS

[01] Session

What the agent sees

A Claude Code session runs the triage prompt. The agent reads the verdict, the transcript and the browser console and network logs, then names the failing request.

Claude Code · testerarmy MCP

The last run of my checkout test failed. Read the transcript and the browser telemetry and tell me whether it's a bug in our app or a problem with the test.

testerarmy(run status)verdict failed

testerarmy(agent transcript)transcript read, credential steps redacted

testerarmy(console and network logs)web run telemetry read

get_guide(debugging)guide loaded

The test agent entered the discount code and pressed Apply, and the network log shows the discount request returned an error, so this is a bug in the app rather than a problem with the test.

No API key
Clients sign in with OAuth 2.1 and dynamic client registration, so there is no API key to rotate.
Web telemetry
Console and network logs are available for web runs once they finish, while mobile runs have the transcript.
Team scoped
Every tool call is scoped to one team, and each tool declares whether it reads, writes or deletes.

[02] Setup

How it works

Setup happens in the coding agent's own MCP configuration, and the first reply before sign-in reads No authorization provided, which is expected until the browser OAuth flow completes.

Add the hosted server

Point Claude Code, Cursor, Codex or VS Code at https://tester.army/mcp, the server TesterArmy hosts, with nothing to configure ahead of time.

Sign in through the browser

On first use the client registers itself and opens a browser for the TesterArmy sign-in, which in Claude Code starts with the /mcp command.

Ask in plain language

Once connected, the agent creates tests, queues runs for a test or a group, reads failures and fetches the built-in guides with get_guide.

Read the setup docsThe docs cover setup for Claude, Cursor, Codex and VS Code, the agent's tools, starting prompts and security.

See it on your own app

START TESTING FOR FREE

[03] Related

Works with

Claude Code, Codex and Cursor each have a setup page for this server with their own commands and session, and the GitHub App tests the pull requests those agents write.

Claude CodeClaude Code creates, runs and reads TesterArmy tests from its own session.CodexCodex creates tests, runs saved groups and reads failures over the hosted MCP server.CursorCursor writes, runs and triages TesterArmy tests without leaving the editor.GitHubThe GitHub App posts checks and comments on every pull request preview.

[04] FAQ

Questions

Through the hosted MCP server, Claude Code, Cursor, Codex and VS Code create saved tests whose steps are written in plain language, queue runs for a test or a group, and poll their status.

The agent reads the run's transcript, browser console logs and network requests, and can fetch a video link that expires in 15 minutes. Console and network telemetry exists for web runs only.

The agent requests a signed upload link and sends the iOS Simulator app or Android build from disk, so the binary never enters the conversation, and runs target the platform's most recent upload.

Claude Desktop, claude.ai, Claude Code, Cursor, Codex and VS Code have setup steps in the docs, any remote MCP client over Streamable HTTP works, and stdio-only clients bridge through mcp-remote.

[05] Start in the editor

Connect a coding agent to TesterArmy

Create a TesterArmy account, point Claude Code, Cursor, Codex or VS Code at https://tester.army/mcp, and sign in over OAuth when the browser opens.