ErrataGuard: AI Board Game Rules & Errata Resolution Agent

Description and use

ErrataGuard uses structured Sanity schemas and the Model Context Protocol to accurately present game rules and errata, eliminating hallucinations and outdated rulings.

Problem

Tabletop gamers struggle with conflicting rulebooks, errata sheets, and developer FAQs, leading to hallucinated or outdated rulings when using generic keyword search or RAG setups.

Outcome

ErrataGuard delivers accurate, source‑attributed rule answers by structuring game rules and errata in Sanity and querying them via MCP, eliminating hallucinations and outdated rulings.

Workflow

• 1. Define custom Sanity schemas for ruleClause and errata, including fields for page numbers, text, version, effective date, and precedence.
• 2. Populate the Sanity Content Lake with rulebook sections and corresponding errata documents.
• 3. Configure Sanity Context to point to the structured dataset and expose it via an MCP endpoint.
• 4. Build an AI agent that streams queries through MCP during chat, reading structured references instead of performing full‑text vector searches.
• 5. When a user asks a rule question, the agent retrieves both the base rule and any overriding errata, displaying them side‑by‑side with precedence ranks.

Evidence

Standard keyword search or generic RAG setups frequently blend rulebook text and errata updates together, resulting in hallucinated or outdated rulings.

Source

ErrataGuard uses structured Sanity schemas and Model Context Protocol (MCP) to guarantee absolute source attribution, ensuring that when an official errata overrides an original rule, both claims surface side‑by‑side with clear precedence ranks.

Source

I structured complex game components inside the Sanity Content Lake using custom schemas (ruleClause and errata), mapping out rulebook sections, version printings, and override priorities.

Source