Cartographer MCP

Builds an AST-based rule system paired with sqlite-vec for semantic memory lookups in agent workflows. This is for orchestrating A-MEM (associative memory) patterns where you need to store and retrieve capability information using both structured rules and vector similarity. You'd reach for this when building agents that need to reason about what actions they can take based on context, using the rule tiers to filter capabilities and the vector database to find relevant precedents or similar situations. The architecture assumes you're working with agent systems that maintain state across interactions and need fast lookups of past decisions or available operations.