Lex

关于

Lex is a TypeScript framework designed to address the critical problem of context loss in AI agent-assisted development. It empowers AI agents with the ability to remember past work sessions, understand architectural constraints, and navigate large codebases efficiently. By offering episodic memory, spatial memory, and architectural policy enforcement, Lex ensures that AI assistants can recall prior actions, respect design boundaries, and provide highly relevant, context-aware assistance, significantly reducing the need for repeated explanations.

主要功能

• Spatial Memory (Atlas Frames) for token-efficient architectural context of module dependencies
• MCP Server Integration exposing 11 tools for seamless interaction with MCP-compatible AI assistants
• Episodic Memory (Frames) for capturing work snapshots with context and next actions
• 0 GitHub stars
• Instructions Generation for maintaining a single source of truth for AI assistant guidance across multiple hosts
• Architectural Policy enforcement to define and validate module boundaries, permissions, and deprecation patterns

使用案例

• Preserve context across work sessions with AI coding assistants and document architectural decisions
• Enable AI agents to recall previous work and navigate large codebases efficiently with token-aware context
• Enforce architectural boundaries via CI to prevent policy violations and onboard new team members with architectural history