Browser testing in Postman Agent Mode
Postman’s new built-in browser in Agent Mode, combined with its Playwright integration, allows developers to generate synchronized UI and API tests from a single recording of user behavior. This innovation closes the gap of "contract drift" between frontend and backend tests, ensuring both test suites reflect actual application interactions. Organizations can now run comprehensive UI and API tests together within their CI pipelines using a single command, making testing more efficient and reliable. Developers can kick off a browser session in Agent Mode, generate Playwright and Postman Collection tests, and integrate the results into their team’s dashboard.
Postman has introduced a significant advancement in testing by integrating a built-in web browser directly into Agent Mode. This new feature, coupled with the existing Postman Playwright integration, enables a unified approach to testing. It allows developers to drive an application within a real browser, record all network traffic, and subsequently generate both UI and API tests from that single, observed behavior. This eliminates the common disconnect where UI and API tests are developed in isolation, often leading to "contract drift" where unspoken changes between frontend and backend go unnoticed until a user reports an issue. This integrated methodology ensures that both testing layers are always in sync with real-world application interactions.
The core problem Agent Mode’s browser tool addresses is the "contract drift" that occurs between UI and API layers. Traditionally, UI tests, like those written with Playwright, might assert only on rendered text, while API tests, often in Postman Collections, are based on an assumed contract. This can lead to a scenario where a frontend change to an undocumented endpoint, or an API returning a new un-schematized field, goes undetected. By generating both test suites from a single recording of browser activity, Postman ensures that any change in the API or UI behavior is reflected across both test sets, preventing silent failures and accelerating discovery of integration issues.
Setting up and using this new feature is streamlined. The browser tool runs directly within Agent Mode itself, requiring no additional installations if Postman is updated to version 12. Users simply open Agent Mode, specify a frontend flow, and observe as the embedded browser executes the task, streaming captured requests into their workspace. The output includes a new Postman Collection containing all frontend-initiated requests and a corresponding Playwright .spec.ts file that mirrors the same flow. This unified capture mechanism ensures that all tests refer to the same observed behavior, eliminating discrepancies that arise from independently maintained test models.
Once tests are generated and reviewed, they can be executed together using a single command via the Postman CLI. This command runs the Playwright suite, captures network traffic during execution, and then validates the observed API calls against the Postman Collection. The result is a single, comprehensive report that covers both the UI and API layers. This integrated testing approach significantly enhances efficiency and reliability within continuous integration pipelines, providing a holistic view of application health and reducing the risk of integration bugs reaching production.
Related articles
Build real agentic apps using CUGA: two dozen working examples on a lightweight harness
CUGA, IBM's open-source Agent Harness, simplifies building agentic applications by handling infrastructure, allowing developers to focus on tools and prompts. It offers pre-assembled components for planning, execution, and state management, significantly reducing development time. CUGA has topped agent benchmarks like AppWorld and WebArena.
OpenAI launches new initiative to help find and patch open source bugs
OpenAI has launched "Patch the Planet," a new initiative in partnership with cybersecurity firm Trail of Bits, to enhance the security of open-source projects. This program aims to assist maintainers in identifying and patching bugs, utilizing OpenAI's AI-powered security tools while reducing the burden on project teams.
PP-OCRv6 on Hugging Face: 50-Language OCR from 1.5M to 34.5M Parameters
Baidu has released PP-OCRv6, an advanced optical character recognition (OCR) model supporting 50 languages. Available on Hugging Face, this version significantly improves accuracy and efficiency across various parameter sizes, from 1.5 million to 34.5 million, marking a substantial leap in multilingual OCR technology.
