TLDR
- Thunders launched its test coverage map at Web Summit 2026 in Lisbon on 9 November.
- The map builds itself from the tests teams already run, with no source code, documentation or setup, and shows where those tests stop on web and native Android and iOS apps.
- Gaps are ranked with an explained points rule, and Thunders AI proposes the tests that close them. Teams approve every new test.
- AI coding assistants such as Claude Code and Cursor can read the map through the Thunders MCP server.
- Available now to Thunders customers. See it at Pavilion 3, Growth area, Booth G1-07.
The map builds itself from the tests teams already run, shows where those tests stop on web and mobile apps, and Thunders AI proposes the missing tests
LISBON, PORTUGAL, 9 November 2026: Thunders, the Paris-based AI software testing company, today launched its test coverage map at Web Summit 2026 in Lisbon. The map shows software teams where their tests stop on web and mobile apps, and Thunders AI proposes the tests that close each gap. Teams approve each new test before it is added. The launch comes as AI coding tools help teams ship code faster than they can check it.
Most teams judge a release by its pass rate. A suite of hundreds of passing tests can still leave a checkout option, a settings page or a payment method untested. Meanwhile, the 2024 DORA report found that a 25% increase in AI adoption was associated with an estimated 7.2% decrease in delivery stability.
Thunders' AI Test Agents already create, run, and debug end-to-end tests. The coverage map closes that loop: it watches what those tests cover, and the test suite grows where coverage is missing. The map:
- Builds itself from existing test runs, with no source code, documentation, or setup. Like a robot vacuum that maps a house while it cleans, Thunders maps an app while its tests run
- Shows where tests stop: unvisited screens that the app links to but no test ever opened, such as Forgot password, Gift cards or API tokens
- Checks every way a control is used, not only whether it was tested. Each interaction counts: a button clicked, a field filled and left empty, a checkbox turned on and off
- Ranks gaps with a clear points rule: key actions such as Save, Delete, or Pay, form fields, and unvisited screens score higher, and each score is explained
- Gives one coverage figure with its scope, such as "up to 22% of interactions tested, 40 of 52 screens measured", traceable to the test runs behind it
- Proposes the tests that close each gap, which the team reviews and approves
The map covers web apps and native Android and iOS apps. Through the Thunders MCP server, AI coding assistants such as Claude Code and Cursor can read the map for a code branch, list the gaps on the screens a change touched, and check the coverage of a single test case.
"A pass rate tells you your tests passed, not where they never went," said Karim Jouini, CEO and co-founder of Thunders. "With the coverage map, our AI Test Agents create and run your tests, see where they stop, and propose the tests that fill the gaps. Your suite grows where it's missing coverage, and your team approves every new test."
The test coverage map is available now to Thunders customers. Thunders is exhibiting at Web Summit 2026 from 9 to 12 November at Pavilion 3, Growth area, Booth G1-07, where Chief Revenue Officer Jean-Francois Bottiau and the Thunders team will demonstrate the coverage map. Karim Jouini is available for remote interviews during the event.
About Thunders
Thunders is an AI-native software testing platform where AI Test Agents build, run and maintain end-to-end tests across web, mobile, API, desktop and enterprise systems such as SAP. Teams write tests in plain English, and the platform keeps them running as applications change, with evidence recorded for every run. Founded in 2025 and headquartered in Paris, Thunders is ISO 27001 and SOC 2 certified and runs as SaaS, hybrid or self-hosted. Learn more at thunders.ai.
