This project is a multi-protocol API simulation and mocking system designed to replace external dependencies during development and testing. It provides an API mocking server, a network traffic proxy, and specialized simulators for language model services and identity providers. The system distinguishes itself through deep AI simulation capabilities, including the emulation of language model providers and Model Context Protocol servers using JSON-RPC 2.0. It supports multi-turn conversational logic, state tracking for AI chat APIs, and the visualization of agent execution through call graphs
Responses is a Python mocking library designed to intercept outgoing HTTP calls made with the Requests library to return predefined simulated responses. It functions as a request verification framework and a network simulation tool, allowing for the verification of application behavior against various API response patterns without making real network calls. The project distinguishes itself through a traffic recorder that captures real network interactions and saves them to files for deterministic replay. It further enables the simulation of complex network scenarios, including the triggering
OHHTTPStubs is an HTTP network mocking library and request interceptor used to simulate server responses for testing application behavior without a live backend. It functions as a network request stubbing tool that captures outgoing network calls and replaces them with predefined mock data, custom status codes, and headers. The library distinguishes itself through its ability to simulate a wide range of network conditions, including artificial latency and connection failures. It supports dynamic response sequencing, allowing the tool to return a series of different responses for the same requ
Nock is a Node.js HTTP mocking library used to intercept outgoing network requests and return predefined responses. It functions as a network request interceptor and HTTP mock server, allowing developers to simulate external API behavior for automated testing without making live network calls. The project includes an API fixture recorder to capture real network interactions and save them as reusable data files for playback. It also features a network latency simulator that introduces artificial delays to outgoing HTTP responses to verify timeout handling and race conditions. The library prov