"OpenAI Gym" matching MCP servers:

MCP server that enables Claude Code to send code and plans for review by OpenAI Codex, returning structured feedback inline.

MCP server that lets coding agents test AI agents. Create YAML test cases, snapshot golden baselines, check for regressions, and generate visual reports all from inside Claude Code or any MCP-compatible tool. Works with LangGraph, CrewAI, OpenAI, Claude, Mistral, and any HTTP API.

Provides a universal bridge to interact with any OpenAI-compatible LLM API (local or cloud), enabling model testing, benchmarking, quality evaluation, and chat operations with performance metrics.

MCP server to mock OpenAI, Anthropic, and Google Gemini responses for testing AI integrations without real API keys or token spend.

An MCP server that allows agents to test and compare LLM prompts across OpenAI and Anthropic models, supporting single tests, side-by-side comparisons, and multi-turn conversations.

Enables agents to lint MCP tool schemas inside a conversation, flagging JSON Schema constructs that break specific clients like Claude, Cursor, Gemini, VS Code Copilot, and OpenAI. It reports per-client compatibility, explains rules, and suggests or applies safe auto-fixes before deployment.

Enables AI agents to fetch daily deterministic constraint challenges and submit JSON answers for immediate, stateless evaluation via MCP.

Enables AI agents to drive SECS/GEM host testing by connecting to equipment, sending SECS messages, running E30 startup conformance reports, and monitoring events and status.

Enables ChatGPT to control Mobile Safari on a real USB-connected iPhone through an OpenAI Secure MCP Tunnel and Appium MCP, including device selection, WDA preparation, and Safari session automation.

Enables AI agents, developers, and CI pipelines to run deterministic, offline localization QA — pseudolocalization, locale linting for separators/dates/placeholders/RTL, and glossary enforcement across 20 locales — without any network access or API key. When an OpenAI-compatible endpoint is configured, it can also translate into multiple locales and lint the results, or otherwise emit structured translation jobs instead of fabricating translations.

An MCP server that performs independent verification of artifacts against criteria using a different AI model lineage (codex, OpenAI, or Gemini) to catch defects that same-family checks might miss.

Panel Review is the guardian agent for AI's highest-stakes coding decisions. Before an agent's riskiest designs, diffs, or commits ship, four frontier models, from OpenAI, Anthropic, Google, and xAI, argue them through and return severity-tagged findings, with review gates the agent cannot silently skip.

Enables AI coding agents to build professional websites using a curated component registry, WCAG-verified design systems and tokens, a human-in-the-loop visual editor, and real-browser Chromium QA that verifies rendered pages and iterates on fixes. It exposes deterministic design decisions, screenshots and evidence-backed quality scores as MCP tools over stdio or Streamable HTTP.

Enables comparison of responses from multiple LLMs (OpenAI, Anthropic, Gemini) to the same prompt, returning a validated divergence score based on sentence embeddings.

Analyzes app screenshots to identify UI/UX issues, compare designs with implementations, and provide actionable fixes using GPT-4o/GPT-5.2 vision capabilities. Supports single/batch analysis, design comparison, and automated report generation for iOS, Android, web, and desktop platforms.

Integrates Factifai testing capabilities with MCP-compatible AI tools, enabling asynchronous creation and retrieval of automated tests using AI models from OpenAI or AWS Bedrock.

Model Context Protocol server that enables API test automation across various AI hosts (Claude, OpenAI, custom) using Postman Collections.

Enables an automated development loop where Google Antigravity IDE performs code generation and fixes while OpenAI Codex CLI acts as an independent QA auditor, with live dashboards, slash commands, and closed-loop remediation.

Autonomously analyzes projects, generates Playwright tests via OpenAI, runs them, classifies failures, and recommends fixes, iterating up to 3 times.

Enables wrapping any command or test run to capture a compact, git-style row diff of SQLite or Cloudflare D1 database changes, then instantly apply inverse SQL to revert them. Also produces short plain-language summaries of the side effects using a local Ollama, LM Studio, vLLM, or OpenAI-compatible model.

Enables running the same prompt(s) across roughly 300 models from Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, Qwen and others, then comparing per-model outputs, latency, token usage, error rates and billed cost side by side. Supports free cost estimates before launching, asynchronous run creation with status polling and cancellation, prepaid x402 credit top-ups, and an optional AI-written comparison summary of the results.

A paid remote MCP server for OpenAI Codex harness, offering structured JSON tools for coding task planning, safe command suggestions, test verdict recording, and run history.

Enables automated visual testing by comparing screenshots against baseline images using OpenAI to verify UI consistency. It generates detailed performance reports in both terminal and Excel formats based on provided user stories and acceptance criteria.

Provides browser automation, AI-powered analysis, visual processing, web scraping, automated test generation, and DevTools analysis capabilities. Supports multiple AI providers (OpenAI, Anthropic, Google, Ollama) for intelligent web interaction and data extraction.

Enables AI-powered document analysis and querying for project documentation using vector embeddings stored in Redis. Supports document upload, context-aware Q\&A, automatic test case generation, and requirements traceability through OpenAI integration.

Enables AI-powered mobile app testing and automation through Appium, using Azure OpenAI to intelligently navigate mobile applications and generate test cases.

MCP server for managing Bugzilla tickets and projects, with a built-in cron job for scheduled polling.