Keploy

Writing integration tests by hand is slow, and hand-made mocks drift away from how real services behave. Keploy takes a different route by listening to the traffic your application already handles and turning it into tests and sandboxes. Because capture happens at the kernel level with eBPF, you do not need to add an SDK or rewrite code to get started. The project also leans into a newer problem: checking code produced by AI coding agents before it ships.

• Records both incoming requests and outgoing calls, so tests reflect real usage
• Generates mocks for databases, queues, caches and external APIs automatically
• Produces human-editable YAML tests from OpenAPI or Postman specs
• Runs the same suites locally, in CI/CD or on Kubernetes
• Keeps the core engine open source under the Apache 2.0 license

Keploy is best for backend teams that ship APIs often and want regression coverage without a large test-writing effort. A practical tip: record a session against a staging environment with realistic data first, then review and trim the generated tests before wiring them into your pipeline. Starting on the free Playground plan is an easy way to see how well generated tests match your service before involving the wider team.