Apidog

AI Agents can generate API tests. But should they write directly into your real API project? Probably not without validation. That is why Apidog CLI now supports a safer workflow for AI-assisted API development: Read → Generate → Validate → Write → Read back → Test With Apidog CLI, Agents and automation workflows can work with APIs, environments, variables, test cases, test scenarios, test suites, reports, import/export, and branches. And before writing complex JSON into Apidog, they can validate it first: apidog cli-schema validate test-case-create --file ./case.json apidog cli-schema validate test-scenario-update --file ./scenario.json The idea is simple: Let the Agent generate. Let the CLI validate. Then write, read back, and test. Explore Apidog CLI: https://lnkd.in/ebc5cUDS #APITesting #AIAgents #DevOps #TestingAutomation #CLI

🤝 A2A Debugger Test and inspect agent-to-agent requests, responses, and communication flows — without leaving Apidog. Debug multi-agent systems from one unified view. ⚡ Give it a try 👉

🤖 AI Agent Debugger Inspect conversation turns, model calls, MCP tool calls, Skill executions, and final outputs — all in one place. Iterate your AI agent faster. ⚡

🚀 2.8.14 Update is here!!! https://lnkd.in/e7nPBiRy ⚡️[Optimization] The free version can now create unlimited sprint branches and API versions. ⚡️[Optimization] Enhanced security for project invitations: invitation links are only valid for invited accounts. ⚡️[Optimization] The Ask AI feature in published documentation now uses a sidebar instead of a search box and supports multi-turn conversations. 🐞[Bug Fix] Fixed an issue where, in the request-first mode, response validation of the endpoint request history could not be viewed. 🐞[Bug Fix] Fixed an issue where the certificate status of custom domains could not be updated in real time. 🐞[Bug Fix] Fixed an issue where the example format in shared documentation was not updated synchronously after modifying parameter types. 🐞[Bug Fix] Fixed an issue where configuring a custom navigation bar in documentation mistakenly displayed internal folders, resulting in 404 errors. 🐞[Bug Fix] Fixed an issue where proxy configuration was not effective when automated tests referenced endpoints across projects.

🚀 2.8.13 Update is here!!! https://lnkd.in/e7nPBiRy ⚡️[Optimization] Support running test scenarios using CLI in sprint branches. ⚡️[Optimization] In post processor scripts, you can use pm.response.socketSession to obtain the certificate information of the endpoint. ⚡️[Optimization] Support exporting the member list of the team. 🐞[Bug Fix] Fixed the issue where the connection was not released after debugging Socket.IO in some cases.

انت كDeveloper، أكيد جربت شعور إنك تخلص الUI وتفضل مستني فتره عشان الbackend يجهزوا API لسه ، أو تضطر تعمل Hardcoded Data وتضيع وقتك في مسحها وتعديلها بعدين ودا بياخد وقت ومجهود طبعا.. يؤسفني اقولك اني فيه طريقه افضل ممكن تستخدمها وهتوفر عليك وقت ومجهود كبير هي ميزة الSmart Mocking في أداة Apidog. ليه الميزة دي مهمه جدا ؟ بيانات حقيقية (مش مجرد Text) بمجرد ما بتحدد نوع البيانات (إيميل، اسم، صورة، سعر)، الأداة بتعمل Generate لبيانات حقيقية متغيرة في كل مرة بتنادي فيها الAPI. Mock Server بضغطة واحدة بيديك URL حقيقي تقدر تستخدمه في كود الFrontend فورا وكأن ال Backend شغال تماماً. لو سميت الحقل city هيرجعلك اسم مدينة، ولو سميته avatar هيرجعلك لينك صورة حقيقي. النتيجة؟ أنت بتخلص شغلك، بتعمل Test للEdge cases، وأول ما الBackend يجهز، كل اللي بتعمله إنك بتغير الBase URL وتتاكد ان كل حاجه شغاله تمام وبس مش لسه هتبدا من الاول تاني الأداة دي حرفياً بتجمع بين (Postman + Swagger + JMeter) في مكان واحد. لو عايز توفر وقتك وتزود إنتاجيتك، لازم تجربها. جربتوا قبل كدة فكرة الSmart Mocking ولا لسه شغالين بالطريقة التقليدية؟ عشان توفر وقت، ده اللينك المباشر لعمل الحساب وتجربة ال Mock Server فورا https://lnkd.in/dG9dJdPe #Apidog #SoftwareDevelopment #Frontend #Backend #SmartMocking #TechTools #Programming Tips

Engineering teams are rushing to make their APIs AI-ready, but most are solving the wrong problem. They focus on adding AI-specific endpoints while ignoring the fundamental issue: contract enforcement. Here's what we're seeing in production: When AI agents interact with poorly governed APIs, the failure rate compounds by 3x compared to human-driven integrations. At scale, this breaks because agents lack the context switching and error recovery capabilities that human developers instinctively apply. The real cost isn't in the initial integration. It's in the maintenance burden when API contracts silently drift. One team we worked with saw their AI workflow incident rate jump from 2 to 27 per week after scaling from 10 to 50 microservices without enforced contracts. This leads to a cascade of issues: AI agents faithfully reproduce errors across your entire system, amplifying what would have been isolated failures into system-wide incidents. Documentation and testing alone won't save you - agents require deterministic, machine-verifiable contracts. The solution isn't more aggressive error handling. It's shifting left to enforce API contracts before deployment. Teams that implement spec-driven validation in their CI pipeline typically see AI integration incidents drop by 45% within the first month. What's your experience with AI agents consuming your APIs? Have you measured the impact of contract drift on your automated workflows?

Really appreciate this post 🙏 Huge thanks to you and your team for sharing your experience with Apidog It always means a lot to see teams actually using the product in their real API workflows. What’s even more exciting for us is seeing more developers and teams from the Middle East and Arabic developer community starting to adopt Apidog. We know how fast the tech ecosystem in this region is growing, and we truly care about supporting developers here with better tools. Feedback from users like you is exactly what helps us keep improving the platform — especially as we continue adding **AI capabilities** to make API development and testing faster. Thanks again for the support, and we’re always listening 🙌🙌

"Can we just run the P0 cases for the payment module?" If you've ever been on a QA or backend team, you know that feeling. Right before release, someone asks this, and suddenly you're diving through folders, manually checking boxes, praying you didn't miss anything new that got added this sprint. Our Growth PM, Sharki SHI wrote about why the traditional "put tests in folders" approach stops working when your test suite hits a certain size—and what to do instead. The basic idea: stop organizing tests by folder. Let rules determine which tests run together. We built this into Apidog 2.8.0+. If your test suite is exploding but your organization strategy is still "more folders," worth a read.👇 #APITesting #QA #EngineeringProductivity