PRODUCT · COVER
CoverOps
See what your tests cover. Know what to test next.
Agents on your running services record which code each automated or manual test actually executed; when the code changes, CoverOps shows what the change touched and which tests last reached it.
- Who it is for
- QA leads who must say what a release was tested against, automation teams with hours-long regression suites, and engineering managers deciding whether a change is safe to merge.
- Last updated
What it is
CoverOps is a runtime test coverage and change impact analysis product. Lightweight agents attach to deployed Java, .NET, Node.js, Python, Go and PHP services and record line coverage per service, file and test while functional, end-to-end and manual tests run. After a commit it lists the tests that last reached the changed lines and posts a pass or fail gate to the pull request.
What it does
Coverage on running services
Line coverage per service, file and test, measured on deployed services during functional, end-to-end and manual testing, not only unit tests.
Change impact analysis
Reads the commits you choose and matches every changed line against what each test executed on its last run.
A changed-code gate
Changed lines no test reached are listed before the change ships, and a pass/fail verdict is posted as a commit status on the pull request.
One flag per service
Java takes one -javaagent flag, Node.js one --require; .NET, Python and Go need a few lines; on Kubernetes one annotation attaches the agents.
Your test tools stay
Playwright and Cypress plugins and a JUnit 5 extension name each test; any other tool sends one header. Manual testing counts through coverage monitoring or ReplayOps sessions.
Cloud or air-gapped
Cloud service on AWS Frankfurt, or your own infrastructure including air-gapped networks. Request and response bodies are never captured.
How it works
- 01
Attach an agent to each service
A startup flag or a Kubernetes annotation; the agent reports which code runs.
- 02
Run your tests as you do today
Automated or manual, with the runner you already use; tests are named by plugin or by one header.
- 03
Read the gate
Coverage per test, the tests that last reached a change, the untested changed lines and a verdict your CI can act on.
Where it sits in the lifecycle
CoverOps measures the Test stage and feeds the release gate: coverage evidence decides go or no-go, and uncovered changed code becomes the next request for TestingOps.
Where it runs
Cloud on AWS in Frankfurt (eu-central-1) or your own infrastructure, including air-gapped networks.
Works with
GitHub, GitLab, Bitbucket and Azure DevOps repositories; Playwright, Cypress, JUnit 5; Postman and k6 via one header; CLI and API for any CI; ReplayOps sessions.
Frequently asked questions
- How is CoverOps different from unit-test coverage?
- Unit-test coverage tells you about unit tests. CoverOps measures deployed services while functional, end-to-end and manual tests run against them, which is where a regression suite actually spends its time.
- Does CoverOps choose and run my tests?
- It lists the tests that last reached the changed code and hands the list to your CI through the CLI or API. Your pipeline runs them; tests with no recorded run are not in the list, so run those too.
- Does it work with manual testing?
- Yes. In coverage monitoring mode every agent keeps adding coverage until you stop it, whether traffic comes from automation or from people clicking through. With ReplayOps each recorded session is joined to the code it ran.
- Which languages are supported?
- Agents exist for Java, .NET, Node.js, Python, Go and PHP. Per-test line attribution works on Java, .NET, Node.js and PHP; Python and Go attribute through call traces.
- Is high coverage proof of quality?
- No. Coverage shows that code ran during a test, not that the test checked it. Use it to find code no test exercised and changes that shipped untested.
- How is it priced?
- By the lines of code you monitor, in four packages from 50K to 1M lines; packages start at $399 a month billed annually. The coverops.ai pricing page lists each package.
Other products in the family
See it on your own stack.
A short call with the team: your setup, your volumes, where the product would run.