Skip to main content
Glama

Memory MCP Server

by inchan
ARCHITECTURE.md1.84 kB
# 시스템 아키텍처 ## 🏗 전체 구조(패키지) ```markdown project/ ├── packages/ │ ├── mcp-server/ # MCP 인터페이스/툴 노출 │ ├── storage-md/ # MD 저장/로드/Front Matter │ ├── index-search/ # FTS/그래프 인덱싱 & 검색 │ ├── assoc-engine/ # 연상(Olima) 엔진 │ └── common/ # 스키마/유틸/로깅 └── docs/ ``` ## 🔄 주요 플로우 (Mermaid 시퀀스) **메모리 생성** ```mermaid sequenceDiagram participant Client as Client(Agent) participant MCP as MCP Server participant Storage as Storage participant Index as Index Client->>MCP: create_memory(content, meta) MCP->>Storage: write Markdown (atomic) MCP->>Index: update FTS + link graph Index-->>MCP: ack MCP-->>Client: { id, path, meta } ``` **메모리 검색** ```mermaid sequenceDiagram participant Client as Client(Agent) participant MCP as MCP Server participant Index as Index participant Assoc as Assoc Engine Client->>MCP: search_memory(query, filter) MCP->>Index: query FTS + link graph Index-->>Assoc: candidates Assoc-->>MCP: reranked candidates MCP-->>Client: results (snippets + links) ``` ## 📁 데이터 모델 (Front Matter 예시) ```markdown --- id: "20250927T103000Z" title: "세션 요약: 인덱스 최적화 아이디어" category: "Resources" # PARA: Projects/Areas/Resources/Archives tags: ["index", "fts5", "performance"] project: "alpha-webapp" # 선택 created: "2025-09-27T10:30:00Z" updated: "2025-09-27T10:30:00Z" links: ["20250926T090000Z", "성능-튜닝-노트"] --- ``` ## 🔐 보안 고려사항 - Local-first, 외부 송출 차단 기본값 - 민감정보 마스킹/필터 - 변경 이력(Git) 및 롤백 전략

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/inchan/memory-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server