Navigate Mind Palace to query entity knowledge about codebase subsystems. Use when understanding architecture, exploring how systems work, or checking for hazards/gotchas before modifying code.
Query keeper entities for codebase knowledge. Each entity guards a subsystem and knows HOW it works, WHY it's designed that way, and WATCH_OUT warnings.
List entities: Read .mind-palace/palace.yaml for entity names
Query an entity: Read .mind-palace/entities/<entityname>.yaml (lowercase)
Available Keepers (22 total):
Each entity YAML contains:
name: MemoryKeeper
slug: memorykeeper # Deterministic ID for cross-agent refs
location: memory
role: "Guardian of hierarchical vector memory..."
topics:
- name: semantic search
how: "Vector-based memory using ChromaDB..."
why: "Finding relevant context requires semantic understanding..."
watch_out: "Attractor basins use specific marker format..."
All palace elements have deterministic slugs for cross-agent communication:
memorykeeper, schedulingkeepermemory-add-message (file-function pattern)memory (file stem)backend (directory)Path format: {region}/{building}/{room} → backend/memory/memory-add-message
Slugs survive regeneration - same codebase produces same slugs.
Each major directory has its own sub-palace:
backend/.mind-palace/ - Python backendadmin-frontend/.mind-palace/ - React admin UItui-frontend/.mind-palace/ - Textual TUImobile-frontend/.mind-palace/ - React Native mobileRoot .mind-palace/ contains shared entities (Keepers). Sub-palaces contain regions/buildings/rooms for their scope.
Before modifying the scheduler:
.mind-palace/entities/schedulingkeeper.yamlUnderstanding memory system:
.mind-palace/entities/memorykeeper.yamlFinding the right entity:
.mind-palace/entities/*.yaml for keyword.mind-palace/palace.yaml - Palace index with entity list.mind-palace/entities/*.yaml - Individual entity definitionsbackend/mind_palace/ - Palace implementation code