Syllable Build
A real build-a-word game — unscramble syllable fragments into a curated real word, with hints and scoring
A playable code-breaking peg game with a real Knuth-style minimax codebreaker mode
Unify Dynamics LLC
Real questions, and the tools your agent calls to answer them.
new_gamesubmit_guessnew_gamecodebreaker_guessprovide_feedbackrender_boardreveal_secretgame_inforules_infoMastermind Play is a fully local, playable code-breaking peg game. new_game starts a game in 'guesser' mode (the engine holds a secret; you guess it) or 'codebreaker' mode (you hold a secret; the engine guesses it), with configurable code_length, colors, allow_repeats, and max_attempts. submit_guess (guesser mode) returns black/white peg feedback and in_progress/win/loss status. codebreaker_guess (codebreaker mode) computes the engine's next guess via real minimax over every code still consistent with feedback so far; provide_feedback records your reported black/white counts. render_board shows the guess history. reveal_secret shows a guesser-mode secret for an end-of-game reveal. game_info reports attempts and remaining candidates. rules_info explains it all. Stateless -- no server-side game storage, the caller passes state between calls (note: because of this, the secret in guesser mode necessarily lives in the state you hold -- avoid printing it before the game ends). Keyless; fully local; zero upstream calls; nothing stored.
Sources: Original implementation of classic Mastermind-style peg-feedback rules and Knuth's well-known 1977 minimax algorithm concept; fully local, zero upstream calls.
The publisher, Unify Dynamics LLC, states that tool inputs are processed in memory and not retained.
8 tools, grouped by job. Every one is read-only or compute-only.
https://mastermindplay.connectorbay.com/mcpA real build-a-word game — unscramble syllable fragments into a curated real word, with hints and scoring
A playable Rock-Paper-Scissors arena with a real pattern-adaptive AI opponent
A real rhyming game — genuine rhymes validated by ending SOUND, not spelling, including classic tricky cases
A real, playable number-guessing game — classic mode and a reverse mode where the engine guesses your number
A playable, stateless Nim engine with a provably optimal nim-sum AI opponent
A real, playable word-search grid game — generate a grid, hunt for hidden words, get hints, or solve it
A fully local, PLAYABLE code-breaking peg game — secret generation, guess submission, real black/white peg feedback, and win/loss-by-attempts detection. Colors are plain numbers (1..N), no trademarked names.
new_game starts a game in one of two modes: guesser (the engine generates a secret and you guess it)
or codebreaker (you hold a secret in your head and the engine tries to guess it). In guesser mode,
submit_guess checks your guess and returns black pegs (right color, right spot) and white pegs (right
color, wrong spot). In codebreaker mode, codebreaker_guess computes the engine's next guess using a real
Knuth-style information-theoretic minimax strategy — it tracks every code still logically consistent with
every piece of feedback so far and always proposes the guess that minimizes the worst case — then
provide_feedback records the black/white counts you report. render_board shows the guess history with
peg symbols. reveal_secret shows a guesser-mode secret for a clean end-of-game reveal. game_info reports
attempts used and (in codebreaker mode) how many codes remain possible. rules_info explains it all.
There is no server-side game storage. Every call takes the full game state (the exact object returned by
the previous call) and returns a new one.
Important for guesser mode: because nothing is stored on the server, the secret code necessarily lives
inside state itself — there's no way around that in a stateless design. Don't print the raw state object
to the player before the game ends; use render_board, which never reveals it.
Not a scripted or random opponent in codebreaker mode — its guesses come from a genuine, provable information-theoretic strategy (restricted to guessing from the still-possible codes, a common effective simplification of Knuth's original five-guess algorithm).
Effective 27 September 2026. Unify Dynamics LLC ("we", "us"), Menlo Park, California.
What Mastermind Play is. Mastermind Play is a connector (an MCP server and REST API) that AI assistants and personal agents call on your behalf to run a local Mastermind-style code-breaking game — generating a secret, checking guesses, computing peg feedback, and choosing the engine's own guesses in codebreaker mode.
What we receive. The arguments of each tool call — the game state you (via your assistant) pass back on every call, your guess or feedback, and the technical headers needed to serve the request (an IP address or a client identifier supplied by the assistant, and a Pro API key if you have one) — plus standard server logs.
Everything is computed locally and nothing is stored. Mastermind Play has no database and no server-side game storage of any kind. The entire game state — including the secret code in guesser mode — lives only in your conversation with your assistant, which passes it back to us on each call and receives an updated copy in the response. We do not persist it, log it, or use it for anything beyond computing that one response. No request is forwarded to any third party — there is no upstream call at all.
What we do not receive or store. We do not receive your name, your conversation with the assistant beyond the bare game-state argument, or anything else about you. Each request is answered independently. Rate-limit counters (a client identifier and timestamps) are held in memory for at most 24 hours. Hosting-provider access logs (IP addresses and request paths, not request bodies) are retained for up to 30 days for security and abuse prevention.
Sub-processors. None — computation is entirely local to this server. Hosting: Vercel (United States regions).
Legal bases and your rights. Our processing is necessary to provide the service you request through your assistant. Because we store no personal data beyond short-lived logs, requests to access, correct or delete personal data are usually satisfied automatically; you can still contact agent@connectorbay.com and we will respond within 30 days. We do not sell or share personal information and do not use your data to train models.
Children. Mastermind Play is a general-audience service operated by the adult using the assistant.
Changes. We will post changes here with a new effective date.
Effective 27 September 2026. These terms govern use of the Mastermind Play connector ("the Service") operated by Unify Dynamics LLC ("we", "us").
Email agent@connectorbay.com. We answer within two business days.
https://mastermindplay.connectorbay.com/mcphttps://mastermindplay.connectorbay.com/openapi.jsonWhen you report a problem, include the question you asked your agent, roughly when, and which agent you used (for example Claude, ChatGPT or Gemini). Never send us passwords, account numbers or ID numbers. Mastermind Play never asks for them.