test-inventory-mcp

An MCP server for test-inventory: AI agents ask which tests a project has, which are skipped and why, without running them.

It reads Playwright and Vitest test files with test-inventory and answers from their source code. Nothing runs, nothing is written, and nothing leaves your machine.

Setup

Node.js 22.12 or newer. The client starts the server with npx and talks to it over standard input and output. Give it the globs of your test files and the framework, as you would to test-inventory's command line.

Claude Code, in the project folder. --scope project writes .mcp.json, for everyone who works on the project:

claude mcp add --scope project test-inventory -- npx -y test-inventory-mcp@0.2 "tests/**/*.spec.ts" --framework playwright

Cursor, in .cursor/mcp.json:

{
  "mcpServers": {
    "test-inventory": {
      "command": "npx",
      "args": ["-y", "test-inventory-mcp@0.2", "tests/**/*.spec.ts", "--framework", "playwright"]
    }
  }
}

VS Code, in .vscode/mcp.json:

{
  "servers": {
    "test-inventory": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "test-inventory-mcp@0.2", "tests/**/*.spec.ts", "--framework", "playwright"]
    }
  }
}

Without globs, as in a configuration shared by every project, the agent passes framework and patterns in each call; the tools say so when they are missing.

The project is the folder given with --root, otherwise the one Claude Code names in CLAUDE_PROJECT_DIR, otherwise the folder the client starts the server in.

Command line

test-inventory-mcp [<glob...> --framework <playwright|vitest>] [--root <dir>]

Option	Meaning
<glob...>	The test files, relative to the project. A tool call can pass others.
--framework <name>	playwright or vitest, for the globs.
--root <dir>	The project. Tool calls can scan folders inside it, never outside.
--help, --version	Print the help or the version.

Tools

Tool	Answers
test_summary	Counts by state, tag and annotation; with groupBy, the counts of each file or tag.
find_tests	The tests that match states, tags, grep, only or withoutReason, with the fields you pick.
get_tests	Every key of up to 20 tests by id, with the scan's diagnostics about them.
scan_diagnostics	The problems the scan found, such as a file that failed to parse, by level, code or path.

Every tool also takes directory, a folder inside the project to scan, and framework and patterns to override the command line. Records have the keys of test-inventory's output format.

Answers are JSON, sent both as structured content and as text. Lists come 50 items at a time (limit takes up to 200), and an answer stays under about 30,000 characters; nextCursor gets the next page. scanErrors counts the problems that may leave tests out, and hint says what to do next when there is something to say.

Security

The server reads the test files its globs match inside the project, and, for Vitest, the files that symbolic links among them point to, as Vitest does. It refuses a directory or pattern that reaches outside the project. It runs no code from the project, writes no files, and opens no network connections.

Answers hold text from the project's files, such as test titles, skip reasons and comments. A client should treat it as content of that project, not as instructions. To report a vulnerability, see SECURITY.md.

Versioning

Tool names, their arguments and their answers are the server's API. Before 1.0, a change that breaks them raises the minor version, so test-inventory-mcp@0.2 in the setup above gets fixes only.

License

MIT