據 AlphaMoat 統計,截至 2026-10,Lancedb Memory 累計下載 11,542 次、獲 7 個星標,按累計下載量在本次參與排名的 10,914 個 AI 技能中排名第 2,383 位,在「AI 智慧體」分類(1,632 個)中排名第 321 位。
列表預設按累計下載量排序,也可按星標排序;詳情頁排名按累計下載量計算。
| 平台 | 累計下載 | 星標 | 評論 |
|---|---|---|---|
| ClawHub ↗ | 6685 | 7 | - |
| SkillHub ↗ | 1.2萬 | 7 | — |
各平台的計數口徑不同,統計時間不一致。
示例包含內容、分類、標籤、重要性和後設資料的儲存、查詢、更新、刪除及統計。依賴 Python 與 LanceDB,並使用本地儲存路徑。語義檢索呼叫的向量與嵌入配置仍需驗證;敏感內容的同意、脫敏、保留和加密控制需額外實現,不能把示例視為已驗證的完整記憶系統。
資料釋出於 2026-10-27
以下是根據技能文件生成的示例,不代表實際執行結果。
2026-09 起統計方法調整:同名技能按作者分別統計,並覆蓋全部技能。圖中虛線為調整前的月份,與之後的資料不直接比較,也不計算跨越調整的增長。
| 月份 | 累計下載 | 環比 | 星標 |
|---|---|---|---|
| 2026-10 | 1.2萬 | — | 7 |
| 2026-09 | 7323 | — | 7 |
| 2026-08 | 6647 | +2.5% | 7 |
| 2026-07 | 6482 | +4.6% | 7 |
| 2026-06 | 6198 | +3.6% | 7 |
| 2026-05 | 5984 | — | 7 |
| 2026-03 | 5234 | — | 7 |
以下是作者釋出的原始檔案,按純文字顯示,本站未做修改或稽核;其中的連結不可點選,命令請核實後再執行。抓取於 2026-10-23。
#!/usr/bin/env python3 """ LanceDB integration for long-term memory management. Provides vector search and semantic memory capabilities. """ import os import json import lancedb from datetime import datetime
from typing import List, Dict, Any, Optional
from pathlib import Path
class LanceMemoryDB:
"""LanceDB wrapper for long-term memory storage and retrieval."""
def __init__(self, db_path: str = "/Users/prerak/clawd/memory/lancedb"):
self.db_path = Path(db_path)
self.db_path.mkdir(parents=True, exist_ok=True)
self.db = lancedb.connect(self.db_path)
# Ensure memory table exists
if "memory" not in self.db.table_names():
self._create_memory_table()
def _create_memory_table(self):
"""Create the memory table with appropriate schema."""
schema = [
{"name": "id", "type": "int", "nullable": False},
{"name": "timestamp", "type": "timestamp", "nullable": False},
{"name": "content", "type": "str", "nullable": False},
{"name": "category", "type": "str", "nullable": True},
{"name": "tags", "type": "str[]", "nullable": True},
{"name": "importance", "type": "int", "nullable": True},
{"name": "metadata", "type": "json", "nullable": True},
]
self.db.create_table("memory", schema=schema)
def add_memory(self, content: str, category: str = "general", tags: List[str] = None,
importance: int = 5, metadata: Dict[str, Any] = None) -> int:
"""Add a new memory entry."""
table = self.db.open_table("memory")
# Get next ID
max_id = table.to_pandas()["id"].max() if len(table) > 0 else 0
new_id = max_id + 1
# Insert new memory
memory_data = {
"id": new_id,
"timestamp": datetime.now(),
"content": content,
"category": category,
"tags": tags or [],
"importance": importance,
"metadata": metadata or {}
}
table.add([memory_data])
return new_id
def search_memories(self, query: str, category: str = None, limit: int = 10) -> List[Dict]:
"""Search memories using vector similarity."""
table = self.db.open_table("memory")
# Build filter
where_clause = []
if category:
where_clause.append(f"category = '{category}'")
filter_expr = " AND ".join(where_clause) if where_clause else None
# Vector search
results = table.vector_search(query).limit(limit).where(filter_expr).to_list()
return results
def get_memories_by_category(self, category: str, limit: int = 50) -> List[Dict]:
"""Get memories by category."""
table = self.db.open_table("memory")
df = table.to_pandas()
filtered = df[df["category"] == category].head(limit)
return filtered.to_dict("records")
def get_memory_by_id(self, memory_id: int) -> Optional[Dict]:
"""Get a specific memory by ID."""
table = self.db.open_table("memory")
df = table.to_pandas()
result = df[df["id"] == memory_id]
return result.to_dict("records")[0] if len(result) > 0 else None
def update_memory(self, memory_id: int, **kwargs) -> bool:
"""Update a memory entry."""
table = self.db.open_table("memory")
valid_fields = ["content", "category", "tags", "importance", "metadata"]
updates = {k: v for k, v in kwargs.items() if k in valid_fields}
if not updates:
return False
# Convert to proper types for LanceDB
if "tags" in updates and isinstance(updates["tags"], list):
updates["tags"] = str(updates["tags"]).replace("'", '"')
table.update(updates, where=f"id = {memory_id}")
return True
def delete_memory(self, memory_id: int) -> bool:
"""Delete a memory entry."""
table = self.db.open_table("memory")
current_count = len(table)
table.delete(f"id = {memory_id}")
return len(table) < current_count
def get_all_categories(self) -> List[str]:
"""Get all unique categories."""
table = self.db.open_table("memory")
df = table.to_pandas()
return df["category"].dropna().unique().tolist()
def get_memory_stats(self) -> Dict[str, Any]:
"""Get statistics about memory storage."""
table = self.db.open_table("memory")
df = table.to_pandas()
return {
"total_memories": len(df),
"categories": len(self.get_all_categories()),
"by_category": df["category"].value_counts().to_dict(),
"date_range": {
"earliest": df["timestamp"].min().isoformat() if len(df) > 0 else None,
"latest": df["timestamp"].max().isoformat() if len(df) > 0 else None
}
}
# Global instance
lancedb_memory = LanceMemoryDB()
def add_memory(content: str, category: str = "general", tags: List[str] = None,
importance: int = 5, metadata: Dict[str, Any] = None) -> int:
"""Add a memory to the LanceDB store."""
return lancedb_memory.add_memory(content, category, tags, importance, metadata)
def search_memories(query: str, category: str = None, limit: int = 10) -> List[Dict]:
"""Search memories using semantic similarity."""
return lancedb_memory.search_memories(query, category, limit)
def get_memories_by_category(category: str, limit: int = 50) -> List[Dict]:
"""Get memories by category."""
return lancedb_memory.get_memories_by_category(category, limit)
def get_memory_stats() -> Dict[str, Any]:
"""Get memory storage statistics."""
return lancedb_memory.get_memory_stats()
# Example usage
if __name__ == "__main__":
# Test the database
print("Testing LanceDB memory integration...")
# Add a test memory
test_id = add_memory(
content="This is a test memory for LanceDB integration",
category="test",
tags=["lancedb", "integration", "test"],
importance=8
)
print(f"Added memory with ID: {test_id}")
# Search for memories
results = search_memories("test memory")
print(f"Search results: {len(results)} memories found")
# Get stats
stats = get_memory_stats()
print(f"Memory stats: {stats}")將開發過程中的經驗、錯誤與修正記錄為 Markdown 檔案,供日後重用並提升為工作區記憶。
用自然語言描述需求,在本機與遠端技能市場尋找技能並取得推薦。
給 AI agent 用的瀏覽器自動化命令列技能。
將使用者明確糾正與自我檢視整理成分層記憶,並依使用情況升級或封存。
規劃 AI 代理主動跟進、工作記憶與回顧流程的工作區設定指南。
指導如何建立與更新有效 Claude 技能的寫作指南。
在安裝第三方技能之前,按清單審查來源、許可權與危險模式的技能。
把盛唐詩人李白的詩學觀、意象系統與口吻,蒸餾成 AI 可呼叫的角色包。
資料月份:2026-10 · 下載/星標數聚合自公開技能註冊庫(ClawHub、SkillHub),按月更新。口徑說明見資料方法。