The Powerduck Blog
Product updates, engineering notes, and guides for designing, debugging, testing, mocking, and serving MCP tools from one OpenAPI spec.
Browse the guidesQuick answer: An agent is surprisingly good at a local change and surprisingly dangerous at a shared one. Give it a helper used across the codebase and ask for a
Read articleYou run the failing command by hand, watch it pass, declare done, and merge. Three months later the same path breaks because nobody turned that manual repro into a test that stays in the suite.
Read articleCode Quality localization, unified sign-in, clearer credit refresh feedback and desktop UI consistency.
Read articleGoogle and GitHub sign-in, resettable AI credits, professional code review and test proposals, connected through AI Chat.
Read articleQuick answer: Mock mappings accumulate like any other code, and over time a test suite builds up dead stubs, mappings no test actually exercises. They look harmless but increase maintenance,
Read articleQuick answer: You do not need to mock an entire API to test the one behavior that matters. A partial mock intercepts only the routes under test and forwards everything
Read articleQuick answer: Built-in linting catches generic mistakes, but every team has conventions that generic rules do not know, naming standards, required headers, versioning rules, and which operations an agent may
Read articleQuick answer: Docs search is a core navigation feature, not an add-on, and the choice between a hosted search index and build-time static search is really a decision about infrastructure,
Read articleQuick answer: Multi-step API workflows require passing a value from one response into the next request, and how that chaining is done determines whether the flow is reliable and typed.
Read articleQuick answer: Pre-request scripts solve a real problem, computing a dynamic value like a signature or a timestamp at send time, but they bury request behavior in imperative code that
Read articleQuick answer: A registry is a discovery channel, not a guarantee of availability. If your agent workflow depends on a server pulled from a public catalog and that entry is
Read articleQuick answer: Pasting a large OpenAPI document into an AI editor does not give the model better API knowledge; it burns the context window on schemas the current task never
Read article