Show HN: Agent VCR – Record, replay, and diff MCP server interactions (github.com)

Agent VCR has been introduced as a revolutionary testing framework tailored for the Model Context Protocol (MCP), allowing developers to record, replay, and diff server-client interactions in a manner reminiscent of VCR technology. This tool provides a solution to common issues faced by MCP server and client developers, such as flaky tests from unreliable live servers, difficulties in error handling testing, and slow CI processes due to real API calls. With Agent VCR, interactions can be recorded once and replayed indefinitely, ensuring that tests run deterministically and offline without the need for a live server, effectively reducing testing time and API quota consumption. The framework supports both Python and TypeScript, featuring a simple installation process and user-friendly commands for recording and replaying interactions. Key technical capabilities include the ability to simulate server errors without altering the actual server, as well as advanced diffing features that help detect breaking changes before deployments. This is especially beneficial for teams managing multiple agents and MCP servers, enabling them to share compatibility checks and maintain consistent performance as the MCP specification evolves. Overall, Agent VCR enhances CI/CD workflows by eliminating flaky tests and accelerating development cycles, significantly contributing to the robustness of AI applications built upon MCP.