Cypress Cloud MCP. Get full E2E test reports via conversation.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Cypress Cloud MCP connects your end-to-end testing platform directly to your AI agent. You can monitor test runs, inspect failed specs for screenshots and error logs, track intermittent failures, or generate detailed performance reports—all without leaving your development environment.
What your AI agents can do
Get instance
Retrieves full details on one specific spec file execution, including videos, screenshots, and error messages.
Get instances
Lists all individual spec files run within a single test job, showing their status and duration.
Get run
Gets comprehensive details for one entire test run, including total tests, passes/fails counts, and commit info.
Find the unique 6-character ID needed to pull project-specific testing data.
Get a summary of recent test executions, including pass/fail counts and commit information for any given project run.
Inspect specific failed spec instances to pull video URLs, screenshots, or detailed error messages.
Generate reports that highlight the slowest tests and track their average duration metrics (p95/avg).
Produce dedicated reports listing flaky tests, showing how often they pass or fail across your codebase.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
Cypress Cloud with 10 Tools
These tools let you manage every aspect of E2E testing, from listing projects to generating comprehensive failure and performance reports.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using Cypress Cloud on Vinkius019d7580get instance
Retrieves full details on one specific spec file execution, including videos, screenshots, and error messages.
019d7580get instances
Lists all individual spec files run within a single test job, showing their status and duration.
019d7580get run
Gets comprehensive details for one entire test run, including total tests, passes/fails counts, and commit info.
019d7580get runs
Lists recent test runs for a project, providing status, duration, and build IDs to track history.
019d7580get tests
Lists every individual test within an instance, returning titles, states (pass/fail), durations, and error messages.
019d7580list projects
Finds all projects under your account to get the unique ID needed for subsequent data extraction.
019d7580report flaky
Generates a report focused on intermittent failures, listing tests that frequently pass and fail.
019d7580report runs
Creates an aggregated summary report of test runs for a specific date range, useful for BI dashboards.
019d7580report slow
Generates a dedicated report showing the slowest tests based on average and p95 duration metrics.
019d7580report tests
Pulls an enterprise-level report containing individual test data, statuses, and error messages across many runs.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Cypress Cloud, then connect any of our 4,800+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,800+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Cypress. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Debugging E2E test failures used to mean hours of log hunting.
Before this, finding the root cause of a failure was painful. You'd get a vague 'Test Failed' status on the dashboard and then spend time copying chunks of logs, pasting them into markdown files, and emailing them to teammates. If the error involved a UI element, you were hoping someone remembered to attach a screenshot from the right tab.
Now, your agent handles the initial triage. You simply ask it about a failure. It goes straight to the source data, pulling up the detailed status, along with video URLs and screenshots for that specific failing test instance. The information is organized, actionable, and immediately available.
Report Flaky Tests Using report_flaky
Previously, tracking intermittent bugs required manually reviewing logs from dozens of runs over months, trying to spot the patterns. You had to maintain a spreadsheet and flag every time 'Checkout Flow' failed just because it popped up in the daily summary.
With `report_flaky`, you get one clean report that aggregates this data for you. It identifies the test name, gives its flake rate percentage, and lists the last date it dropped. You stop tracking failures; you start fixing them.
What you can do with this MCP connector
This connection gives you total control over your quality assurance lifecycle using plain conversation. Instead of navigating through Cypress Cloud dashboards to find a specific failure log, you just ask your AI client about it. You can request a list of recent runs and get immediate status summaries, or dive deep into one failing test case to pull up the video and screenshot evidence right away.
It's also useful for finding performance bottlenecks; you can generate slow test reports to check average duration metrics across your CI/CD pipeline. Furthermore, if you need an audit trail of intermittent bugs, the MCP identifies flaky tests and tracks their last failure dates. Because Vinkius handles all data flow through a zero-trust proxy, your credentials never sit on disk, keeping everything secure while letting your agent chain this test data with other services in your workflow.
019d7580-c770-70dc-9094-96c2ccda2db3 
How Cypress Cloud MCP Works
- 1 Subscribe to the MCP and supply your Cypress Cloud Enterprise API Key along with your Project ID.
- 2 Connect this MCP through any compatible client, like Cursor or Claude, giving your agent access to the test data endpoints.
- 3 Ask your agent a question—like 'What failed in the last run?'—and it uses the tools to retrieve and summarize the results.
The bottom line is you get instant answers about your E2E test quality, without opening the Cypress UI.
Who Is Cypress Cloud MCP For?
This is for QA Engineers who hate clicking through dashboards just to find one error log. It's also for DevOps teams needing reliable pipeline data and developers who want test results right in their IDE.
Uses the MCP to query recent runs, quickly filtering out passing tests until they find a failing spec instance. They then pull error messages and screenshots for immediate debugging.
Runs reports to check pipeline reliability, specifically using the capability that identifies flaky tests across multiple commits.
Asks the agent for test results and error logs directly from their workspace instead of switching context to a separate dashboard tool.
What Changes When You Connect
- Stop manually checking run logs. You can use
get_runsto list recent executions and immediately see the status, commit info, and total pass/fail counts for a project. - When something breaks, you don't waste time hunting through videos. Use
get_instanceto pull up detailed error messages and screenshots right from your agent chat. - Improve CI reliability by running
report_flaky. This tool identifies intermittent bugs, tracking failure rates and last known dates for those hard-to-catch issues. - Track performance degradation using
report_slow. This gives you concrete data on which tests are slowing down your build process based on p95 metrics. - Need historical context? The combination of
get_runsandreport_testslets you gather massive amounts of run summary and granular test data for auditing or BI dashboards.
Real-World Use Cases
Debugging a recent failure
A QA Engineer sees a red status on the dashboard. Instead of clicking through, they ask the agent to find all failed specs in that run using get_instances. The agent responds with links and error messages for the top three failures.
Auditing quarterly reliability
An Engineering Manager needs a report showing process quality across Q2. They ask the agent to generate an enterprise run summary using report_runs for the last quarter, receiving data formatted for a board presentation.
Tackling 'ghost' bugs
A developer suspects a test is flaky but can't prove it. They ask the agent to run report_flaky and get confirmation that the 'User Profile Update' test has failed three times this month, narrowing down the root cause.
Reviewing pipeline bottlenecks
A DevOps team member notices build time creep. They ask the agent to run report_slow and get a list of tests that have exceeded an average duration threshold, allowing them to prioritize refactoring.
The Tradeoffs
Assuming one report covers everything
Asking the agent for 'the full test results' without specifying if you need a run summary or individual test data.
→
If you need overall metrics, use get_run. If you need historical context across dates, ask for a specific range using report_runs.
Ignoring the project ID
Trying to pull test data without knowing which project container holds the specs.
→
Always start by calling list_projects. This gives you the unique 6-character ID needed before running any other reporting tool.
Confusing run status with spec status
Seeing that a whole test run failed, but not knowing which specific spec file caused it.
→
First, use get_runs to see the failure. Then, narrow down by calling get_instances on that run ID to isolate the failing spec.
When It Fits, When It Doesn't
Use this MCP if your core problem is deep visibility into test quality and performance metrics—specifically identifying why a test failed (the screenshot/error log) or if it fails intermittently. If you only need simple build status checks (e.g., 'Did the last commit pass?'), get_runs is enough, but if you need to debug, use this MCP. Don't use this if you are trying to manage user accounts or payment processing; those require different tooling altogether.
Common Questions About Cypress Cloud MCP
How do I find my project ID using list_projects? +
You call list_projects. This tool returns all your organizational projects, providing the exact 6-character IDs you need to reference in subsequent report calls.
What is the difference between get_runs and get_instances? +
get_runs gives a high-level summary of an entire test job (pass/fail counts, total duration). get_instances drills down into that run to list every single spec file execution within it.
Can I get performance data on slow tests with report_slow? +
Yes. report_slow generates a dedicated report showing the slowest tests, quantifying their issue using average (avg) and 95th percentile (p95) duration metrics.
Does get_tests show me individual failures? +
Yes. get_tests lists every single test title within an instance, showing its specific state (passed/failed/skipped), along with error messages if it failed.
When I use get_instance, what detailed debugging artifacts can I retrieve for a failed test? +
The tool gives you full debug context. You don't just get an error message; you also receive screenshots and video URLs specific to that failing spec instance. This lets you review exactly what the browser saw when the failure happened.
If I need a historical view of test results, how do I use report_runs? +
You must provide a start date for this tool. It generates an enterprise run summary report, giving you aggregated data across multiple runs. This is ideal for feeding into BI dashboards or long-term auditing.
What metrics does report_flaky use to identify unreliable tests? +
It identifies intermittent failures by analyzing historical pass/fail cycles. The output provides the test name, a calculated flake rate percentage, and crucially, lists the last dates that specific test failed.
Is report_tests better for overall reporting than just listing results with get_tests? +
Yes, report_tests provides comprehensive, individual test-level data. While get_tests lists current state and errors, this report fetches the full historical status and error messages for a much broader dataset.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.