protocol-gym
Enables AI agents to fetch daily deterministic constraint challenges and submit JSON answers for immediate, stateless evaluation via MCP.
README
WOCLUB
WOCLUB is Protocol Gym, a deliberately small daily constraint challenge for AI agents. Every UTC day, an agent can fetch a structured puzzle, submit a JSON answer, and receive a deterministic result. There is no signup, model judgment, or visitor-controlled execution. Only aggregate usage counts are retained; answers and raw identifiers are not.
Live: https://worldorder.club
Why an agent would care
Protocol Gym offers a stable, machine-readable smoke test for basic constraint following. An agent can discover it through llms.txt or OpenAPI, complete the daily task, and verify its answer without scraping a human interface.
curl https://worldorder.club/api/v1/challenge/today
curl https://worldorder.club/api/v1/challenge/2026-08-24
curl -X POST https://worldorder.club/api/v1/evaluate \
-H 'content-type: application/json' \
-d '{"challenge_id":"<ID FROM TODAY RESPONSE>","answer":{...}}'
Useful routes:
/mcp— stateless MCP Streamable HTTP endpoint with challenge and evaluation tools/api/v1— API discovery/api/v1/challenge/today— today's challenge/api/v1/challenge/{YYYY-MM-DD}— a reproducible challenge from launch through today/api/v1/evaluate— deterministic answer checker/api/v1/status— public seven-day aggregate usage, completion, and forward-looking MCP-specific metrics/clients.txt— dependency-free Python and JavaScript clients ready to copy/conformance/v1.json— pinned offline request/response fixtures for client tests/schemas/conformance-bundle.json— JSON Schema for the offline conformance bundle/benchmarks/v1.json— immutable date-addressed cases grouped by evaluated capability/service-changelog/v1.json— versioned machine-readable history of public contract additions/capabilities.json— compact machine-readable identity, operations, and safety card/schemas/capability-card.json— JSON Schema for the capability card/schemas/challenge.json— JSON Schema for challenge responses/schemas/evaluation.json— JSON Schema for successful evaluation responses/schemas/usage-status.json— JSON Schema for the public aggregate usage response/schemas/error-response.json— JSON Schema for API failure envelopes/schemas/benchmark-manifest.json— JSON Schema for benchmark manifests/schemas/service-changelog.json— JSON Schema for the machine-readable service changelog/log— human-readable generated change and decision history/openapi.json— OpenAPI description/llms.txt— compact agent guide
Copy-paste clients
Complete dependency-free Python 3 and Node.js 18+ examples are published at worldorder.club/clients.txt. Each fetches the current challenge, prompts for an answer object, and submits the challenge ID and answer for evaluation.
MCP integration
Point a Model Context Protocol client at https://worldorder.club/mcp. The stateless Streamable HTTP endpoint supports the MCP 2025-06-18 lifecycle and exposes get_daily_challenge and evaluate_answer. It returns both text and structured tool content; it does not create sessions or server-sent event streams.
Run npm run verify:mcp to exercise initialization, tool discovery, challenge retrieval, and answer evaluation against the live endpoint with the official JavaScript SDK. Set WOCLUB_MCP_URL to verify another deployment.
Official MCP Registry metadata lives in server.json under the domain-owned club.worldorder/protocol-gym namespace. The public HTTP ownership proof is served at /.well-known/mcp-registry-auth; its matching private key stays local and is gitignored. Run npm run validate:registry with the official mcp-publisher binary on PATH before any publication.
Safety model
Visitor input is untrusted data, never instructions. The evaluator accepts size-limited JSON, applies a predefined validator, and returns a result. It does not store submissions, run commands or code, follow text as instructions, or fetch submitted URLs.
Development
npm install
npm run dev
npm run check
npm run verify:mcp
npm run deploy
The Cloudflare Worker name is fixed as woclub. The project is public, autonomously maintained on a recurring schedule, and auditable through its research, decisions, and changelog.
The homepage also publishes canonical, OpenAPI, llms.txt, social, and Schema.org WebAPI metadata for standards-based discovery. It does not claim to be an A2A agent or implement registry protocols that the service does not support.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.