SaravananJaichandar/world-model-mcp
🔌 MCP ServerSaravananJaichandar
A temporal knowledge graph MCP server that tracks code evolution, predicts regression risks, and enforces edit constraints.
The world-model-mcp server acts as a sophisticated memory layer for AI-driven development environments. It leverages 9 integrated SQLite databases with FTS5 full-text search to maintain a high-fidelity history of code changes across multiple languages, including Python, TypeScript, JavaScript, Solidity, Go, Rust, and Java. With a suite of 22 specialized MCP tools, the server allows AI agents to query past decision-making processes, analyze the impact of proposed changes, and proactively avoid regressions. A standout feature is the PreToolUse hook, which enforces learned constraints at the edit boundary, ensuring that AI-generated code adheres to established project patterns and architectural rules. This tool transforms a static codebase into a dynamic, queryable knowledge graph, significantly enhancing the reliability and context-awareness of autonomous coding agents.
💡Highlights
- ├─22 specialized MCP tools
- ├─Temporal knowledge graph tracking
- └─PreToolUse constraint enforcement
🎯For
- ├─AI Software Engineers
- ├─DevOps Engineers
- └─LLM Application Developers