Natural-Language-to-Game-Rules Synthesizer

You are the gameplay tools architect for [INSERT GAME TITLE]; design a system that turns natural-language rules into verified, balanced gameplay logic. Non-programmer creators describe modes such as capture-the-point or survival waves; the system compiles intent into safe, testable rules for [INSERT ENGINE SCRIPTING LAYER]. Module 1 - Gameplay design: define the rule DSL's conceptual scope (win conditions, scoring, spawns, modifiers), supported genres, and the creator-facing explanation model. Module 2 - Generation pipeline: specify intent parsing, DSL synthesis, deterministic compilation to engine events, ambiguity clarification, and example-driven repair when tests fail. Include a sandboxed runtime with resource limits and static-analysis gates blocking access to economy, moderation, and cross-game APIs. Provide an LLM-assisted balance layer that simulates thousands of bot matches and proposes parameter changes before publish. Define how generated rules compose with official matchmaking and anti-cheat systems. Explain how creators diff, version, and roll back rules after release. Specify multilingual parsing strategy and locale-specific numeric and time handling. Module 3 - Data structures: provide schemas for rule documents, DSL ASTs, compilation diagnostics, simulation reports, and published versions. Module 4 - Testing: define worked-example equivalence checks, property tests for termination and scoring bounds, bot tournaments, and malicious-prompt suites. Module 5 - Operations: cover DSL versioning and migration, a template marketplace, outcome telemetry, hotfix channels, and creator analytics. Provide a text architecture diagram, three end-to-end NL-to-rule examples, risk register, and KPIs for first-publish success rate. List assumptions about the engine networking model and target creator skill level. Begin with a one-page design summary, then deliver each module one at a time for approval. ================