The Powerduck Blog
Product updates, engineering notes, and guides for designing, debugging, testing, mocking, and serving MCP tools from one OpenAPI spec.
Browse the guidesMost MCP servers load everything eagerly, dump huge schemas into the system prompt, and make the agent pay for tools it will never call. The fix is boring: lazy, narrow, and spec-driven.
Read articleWhen the model writes both the handler and the test, green CI proves almost nothing. Here's the cheap structural fix that catches the failure mode.
Read articleAn AI review bot that reads a diff the AI itself wrote is the author reading their own essay out loud to a friend who already agrees. The only useful review comes from a signal the code didn't generate.
Read articleThe default move when an agent gets something wrong is to paste more files into the context. But more files means more noise, and noise is what made it wrong in the first place.
Read articleA clean architecture diagram can hide a broken transition. Trace one failure boundary, verify it in code, and turn it into an API scenario test.
Read articleA failing request needs its environment, caller, and starting state. Preserve those conditions when handing an API investigation to AI.
Read articleA create-read-check scenario connects an API contract, MCP tools, and observable outcomes. Here is how to make completion something you can verify.
Read articleA practical way to separate request format errors from response negotiation problems, with cURL examples and a four-case regression checklist.
Read articleA practical way to compare a working cURL request with a failing browser request, inspect preflight headers, and verify both success and error responses.
Read articleMCP connects agents to tools; A2A connects agents to other agents. Debugging A2A means pinning the incompatible 1.0/0.3 wire version, reading the Agent Card, walking the task lifecycle, streaming events over SSE, and pinning the JWKS out of band.
Read articleYou already have a working REST API. Hand-writing and maintaining a second, parallel set of AI tool wrappers is pure transcription that drifts. Generate MCP tools from your OpenAPI contract, start read-only, and keep authorization where it belongs - in your API.
Read articleYour AI coding assistant keeps using pageSize instead of limit and data.items instead of data.records because it guesses from pasted docs. Give it a local OpenAPI spec it can query over MCP, then verify multi-step scenarios against real requests and mock the missing dependency.
Read article