Model Context Protocol (MCP)

Model Context Protocol (MCP) is a standard for LLMs to interact with development environments. MCP makes LLM-powered applications such as Claude, Claude Code, and editors like Zed, Cursor, or VS Code interoperable with external services and data sources.

Tuist hosts a server-side MCP endpoint at https://tuist.dev/mcp. By connecting your MCP client to it, AI agents can access your Tuist project data, including test insights, flaky test analysis, and more.

MCP vs Skills

MCP and Skills can overlap in what they do. Given the current overlap between the two, choose one approach per workflow and use it consistently (either MCP or skills) instead of mixing both in the same flow.

Configuration

Add https://tuist.dev/mcp as a remote MCP server in your client. Authentication happens through OAuth automatically.

Claude Code

Run:

claude mcp add --transport http tuist https://tuist.dev/mcp

Claude Desktop

Open Settings → Connectors → Add custom connector, then set:

• Name: tuist
• URL: https://tuist.dev/mcp

Complete OAuth in the browser when prompted.

Cursor

Open Cursor Settings → Tools & Integrations → MCP Tools and add:

• Name: tuist
• URL: https://tuist.dev/mcp

VS Code

Use Command Palette → MCP: Add Server, then configure an HTTP server with:

• Name: tuist
• URL: https://tuist.dev/mcp

Zed

Open Agent panel → Settings → Add Custom Server, then set:

• Name: tuist
• URL: https://tuist.dev/mcp

Capabilities

Tools

The following tools are available through the Tuist MCP server:

Tool	Description	Required parameters
list_projects	List all projects accessible to the authenticated user.	None
list_test_cases	List test cases for a project (supports filters like flaky).	account_handle, project_handle
get_test_case	Get detailed metrics for a test case including reliability rate, flakiness rate, and run counts.	test_case_id or identifier + account_handle + project_handle
get_test_run	Get detailed metrics for a test run.	test_run_id
get_test_case_run	Get failure details and repetitions for a specific test case run.	test_case_run_id

list_projects

Returns all projects the authenticated user has access to, including each project's id, name, account_handle, and full_handle.

list_test_cases

Returns test cases for a given project. Supports pagination through page and page_size, and optional filters such as flaky=true, quarantined=true, module_name, suite_name, and name.

get_test_case

Returns detailed metrics for a specific test case: reliability rate (success percentage), flakiness rate (over the last 30 days), total and failed run counts, last status, and average duration. Accepts either a test_case_id (UUID) or an identifier in Module/Suite/TestCase or Module/TestCase format together with account_handle and project_handle.

get_test_run

Returns detailed test run context: status, duration, CI metadata, and aggregate counts (total/failed/flaky). Use get_test_case_run to drill into individual failures or crashes.

get_test_case_run

Returns the full details of a specific test case run, including failure messages with file paths and line numbers, repetition results, git branch, commit SHA, and whether the run was on CI.

Prompts

Prompt	Description
fix_flaky_test	Guides you through fixing a flaky test by analyzing failure patterns, identifying the root cause, and applying a targeted correction.

The fix_flaky_test prompt accepts optional arguments:

• account_handle and project_handle: scope the investigation to a specific project.
• test_case_id: jump directly to investigating a specific flaky test case by UUID (list_test_cases[].id) or identifier (Module/Suite/TestCase or Module/TestCase).

When invoked without arguments, the prompt guides the agent through discovering flaky tests across all accessible projects.