Persistent memory + RAG + User Profile cho AI agents. #1 LongMemEval benchmark. Thay thế L1/L2 memory thủ công bằng API cloud chuẩn.
Scanned 9/9/2026
Install to Claude Code
npx -y skills add yanacuti1121/Yana-AI --skill supermemory --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Supermemory?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/yanacuti1121-supermemory-yana-ai)More formats (shields.io, HTML) on the badges page.
---
name: supermemory
description: Persistent memory + RAG + User Profile cho AI agents. #1 LongMemEval benchmark. Thay thế L1/L2 memory thủ công bằng API cloud chuẩn.
triggers:
- supermemory
- persistent memory
- nhớ qua session
- AI memory
- long-term memory
- user profile AI
- RAG search
- tìm kiếm ngữ nghĩa
- agent memory
---
# supermemory — Memory Infrastructure for AI
**Source**: github.com/supermemoryai/supermemory ⭐24k
**API**: api.supermemory.ai
**MCP**: mcp.supermemory.ai/mcp
**Benchmark**: #1 LongMemEval · LoCoMo · ConvoMem
## Setup nhanh (MCP — không cần cài local)
```bash
# Thêm MCP server vào Claude Code
npx -y install-mcp@latest https://mcp.supermemory.ai/mcp --client claude --oauth=yes
```
Hoặc thêm thủ công vào `~/.claude/settings.json`:
```json
{
"mcpServers": {
"supermemory": {
"url": "https://mcp.supermemory.ai/mcp",
"headers": {
"Authorization": "Bearer sm_YOUR_API_KEY"
}
}
}
}
```
Lấy API key: [app.supermemory.ai](https://app.supermemory.ai)
## 3 khả năng chính
### 1. Memory API — tự học từ conversation
```python
from supermemory import Supermemory
client = Supermemory(api_key=os.environ["SUPERMEMORY_API_KEY"])
# Lưu memory sau mỗi conversation
client.add(
content="Anh Tâm thích Rust hơn Java, đang dùng Cloud Shell",
container_tag="tam_profile",
metadata={"type": "preference", "date": "2026-06-03"}
)
# Tìm context liên quan
context = client.profile(
container_tag="tam_profile",
query="thói quen làm việc"
)
print(context.profile) # user profile tự generate
print(context.memories) # relevant memories
```
### 2. RAG — tìm kiếm ngữ nghĩa
```python
# Upload tài liệu
client.add(content=open("guide.md").read(), container_tag="yamtam_docs")
# Search hybrid (semantic + keyword)
results = client.search(query="cách cài headroom", container_tag="yamtam_docs")
for r in results.results:
print(r.content[:200])
```
### 3. User Profile — 50ms latency
```python
# Profile tự build từ memories, cập nhật liên tục
profile = client.profile(container_tag="tam_profile")
# → stable facts + recent activity combined
```
## Tích hợp với YAMTAM L1/L2
| YAMTAM hiện tại | Supermemory thay thế |
|---|---|
| `core/scripts/add-fact.sh` | `client.add(content, container_tag)` |
| L1 INDEX.md manual | `client.profile()` tự generate |
| grep search trong .md files | `client.search()` semantic |
| Session context bị mất | Persistent across sessions |
## Connectors (auto-sync)
```python
# Kết nối Google Drive → tự sync
client.connectors.create(type="google_drive", credentials=...)
# GitHub repo sync
client.connectors.create(type="github", repo="yana-ai")
```
Supported: Google Drive · Gmail · Notion · OneDrive · GitHub
## Install SDK
```bash
pip install supermemory # Python
npm install supermemory # TypeScript/Node
```
## Liên quan YAMTAM
- Thay thế `memory/` manual markdown files
- Persistent context qua nhiều session (không mất khi compaction)
- Search ngữ nghĩa qua toàn bộ project history
- User profile Tâm tự cập nhật theo thời gian
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!