majiayu000/spellbook / skills/xiaohongshu
xiaohongshu
Verified生成小红书文案和配图。当用户需要写小红书、生成社交媒体文案、小红书运营内容时使用
| File | Kind | Size |
|---|---|---|
| .env.example | other | 241 B |
| SKILL.md | skill md | 12.9 KB |
| database/feeds.json | config | 99 B |
| database/summary.md | markdown | 885 B |
| reference/extended.md | markdown | 6.5 KB |
| requirements.txt | manifest | 9 B |
| scripts/feed_database.py | script | 11.4 KB |
| scripts/generate_image.pyexecutable | script | 7.6 KB |
| styles/contrast-impact.md | markdown | 1.3 KB |
| styles/cute-illustration.md | markdown | 1.1 KB |
| styles/glass-card.md | markdown | 1.0 KB |
| styles/photo-realistic.md | markdown | 1.1 KB |
| styles/text-highlight.md | markdown | 1.3 KB |
| tag-database.md | markdown | 5.3 KB |
scripts/feed_database.py
View on GitHub#!/usr/bin/env python3
"""
小红书高赞笔记本地数据库
支持:添加、查询、分析、导出
"""
import json
import sys
import os
from pathlib import Path
from datetime import datetime
from collections import Counter
DB_DIR = Path.cwd() / "database"
DB_FILE = DB_DIR / "feeds.json"
SUMMARY_FILE = DB_DIR / "summary.md"
def _load_db() -> dict:
if DB_FILE.exists():
with open(DB_FILE, "r", encoding="utf-8") as f:
return json.load(f)
return {"feeds": [], "meta": {"created_at": datetime.now().isoformat(), "version": 1}}
def _save_db(db: dict):
DB_DIR.mkdir(parents=True, exist_ok=True)
with open(DB_FILE, "w", encoding="utf-8") as f:
json.dump(db, f, ensure_ascii=False, indent=2)
def add(feed_json: str):
"""添加一条笔记到数据库,输入为 JSON 字符串"""
feed = json.loads(feed_json)
required = ["feed_id", "title"]
for key in required:
if key not in feed:
print(json.dumps({"success": False, "error": f"缺少必填字段: {key}"}))
sys.exit(1)
db = _load_db()
# 去重:同 feed_id 则更新
existing_ids = {f["feed_id"]: i for i, f in enumerate(db["feeds"])}
feed["updated_at"] = datetime.now().isoformat()
if feed["feed_id"] in existing_ids:
idx = existing_ids[feed["feed_id"]]
feed["created_at"] = db["feeds"][idx].get("created_at", feed["updated_at"])
db["feeds"][idx] = feed
action = "updated"
else:
feed["created_at"] = feed["updated_at"]
db["feeds"].append(feed)
action = "added"
_save_db(db)
print(json.dumps({"success": True, "action": action, "feed_id": feed["feed_id"], "total": len(db["feeds"])}))
def list_feeds(domain: str = None, min_likes: int = 0, limit: int = 50):
"""列出数据库中的笔记"""
db = _load_db()
feeds = db["feeds"]
if domain:
feeds = [f for f in feeds if f.get("domain") == domain]
if min_likes > 0:
feeds = [f for f in feeds if f.get("likes", 0) >= min_likes]
# 按赞数降序
feeds.sort(key=lambda x: x.get("likes", 0), reverse=True)
feeds = feeds[:limit]
result = []
for f in feeds:
result.append({
"feed_id": f["feed_id"],
"title": f.get("title", ""),
"author": f.get("author", ""),
"likes": f.get("likes", 0),
"favorites": f.get("favorites", 0),
"comments": f.get("comments", 0),
"domain": f.get("domain", ""),
"created_at": f.get("created_at", ""),
})
print(json.dumps({"total": len(db["feeds"]), "filtered": len(result), "feeds": result}, ensure_ascii=False, indent=2))
def get(feed_id: str):
"""获取单条笔记完整数据"""
db = _load_db()
for f in db["feeds"]:
if f["feed_id"] == feed_id:
print(json.dumps(f, ensure_ascii=False, indent=2))
return
print(json.dumps({"error": f"未找到 feed_id={feed_id}"}))
sys.exit(1)
def analyze():
"""分析数据库中所有笔记,生成洞察摘要"""
db = _load_db()
feeds = db["feeds"]
if not feeds:
print(json.dumps({"error": "数据库为空"}))
return
total = len(feeds)
likes_list = [f.get("likes", 0) for f in feeds]
favs_list = [f.get("favorites", 0) for f in feeds]
comments_list = [f.get("comments", 0) for f in feeds]
# 互动数据统计
avg_likes = sum(likes_list) / total
avg_favs = sum(favs_list) / total
avg_comments = sum(comments_list) / total
# 收藏/赞比
fav_like_ratios = []
for f in feeds:
likes = f.get("likes", 0)
if likes > 0:
fav_like_ratios.append(f.get("favorites", 0) / likes)
avg_fav_ratio = sum(fav_like_ratios) / len(fav_like_ratios) if fav_like_ratios else 0
# 领域分布
domain_counter = Counter(f.get("domain", "未分类") for f in feeds)
# 标题类型分布
title_type_counter = Counter()
for f in feeds:
analysis = f.get("analysis", {})
if analysis.get("title_type"):
title_type_counter[analysis["title_type"]] += 1
# 配图风格分布
image_style_counter = Counter()
for f in feeds:
analysis = f.get("analysis", {})
if analysis.get("image_style"):
image_style_counter[analysis["image_style"]] += 1
# 内容类型分布
content_type_counter = Counter()
for f in feeds:
analysis = f.get("analysis", {})
if analysis.get("content_type"):
content_type_counter[analysis["content_type"]] += 1
# 提取高频关键要素
all_elements = []
for f in feeds:
analysis = f.get("analysis", {})
all_elements.extend(analysis.get("key_elements", []))
element_counter = Counter(all_elements)
# 提取高频标签
all_tags = []
for f in feeds:
analysis = f.get("analysis", {})
all_tags.extend(analysis.get("tags_used", []))
tag_counter = Counter(all_tags)
# TOP 5 高赞
top5 = sorted(feeds, key=lambda x: x.get("likes", 0), reverse=True)[:5]
insights = {
"total_feeds": total,
"engagement": {
"avg_likes": round(avg_likes, 1),
"avg_favorites": round(avg_favs, 1),
"avg_comments": round(avg_comments, 1),
"avg_fav_like_ratio": round(avg_fav_ratio, 2),
},
"domains": dict(domain_counter.most_common()),
"title_types": dict(title_type_counter.most_common()),
"image_styles": dict(image_style_counter.most_common()),
"content_types": dict(content_type_counter.most_common()),
"top_elements": dict(element_counter.most_common(10)),
"top_tags": dict(tag_counter.most_common(15)),
"top5_feeds": [{"title": f.get("title"), "likes": f.get("likes", 0), "author": f.get("author", "")} for f in top5],
"analyzed_at": datetime.now().isoformat(),
}
print(json.dumps(insights, ensure_ascii=False, indent=2))
# 同时生成 markdown 摘要
_generate_summary(insights, feeds)
def _generate_summary(insights: dict, feeds: list):
"""生成 markdown 格式的分析摘要"""
DB_DIR.mkdir(parents=True, exist_ok=True)
lines = [
"# 小红书高赞笔记分析摘要",
"",
f"> 数据库共 {insights['total_feeds']} 篇笔记 | 更新时间:{datetime.now().strftime('%Y-%m-%d %H:%M')}",
"",
"---",
"",
"## 互动数据均值",
"",
f"| 指标 | 均值 |",
f"|------|------|",
f"| 赞 | {insights['engagement']['avg_likes']} |",
f"| 收藏 | {insights['engagement']['avg_favorites']} |",
f"| 评论 | {insights['engagement']['avg_comments']} |",
f"| 收藏/赞比 | {insights['engagement']['avg_fav_like_ratio']} |",
"",
]
if insights["domains"]:
lines.extend(["## 领域分布", ""])
for domain, count in insights["domains"].items():
lines.append(f"- {domain}: {count}篇")
lines.append("")
if insights["title_types"]:
lines.extend(["## 标题类型分布", ""])
for ttype, count in insights["title_types"].items():
lines.append(f"- {ttype}: {count}篇")
lines.append("")
if insights["image_styles"]:
lines.extend(["## 配图风格分布", ""])
for style, count in insights["image_styles"].items():
lines.append(f"- {style}: {count}篇")
lines.append("")
if insights["content_types"]:
lines.extend(["## 内容类型分布", ""])
for ctype, count in insights["content_types"].items():
lines.append(f"- {ctype}: {count}篇")
lines.append("")
if insights["top_elements"]:
lines.extend(["## 高频成功要素 TOP 10", ""])
for elem, count in insights["top_elements"].items():
lines.append(f"- {elem} ({count}次)")
lines.append("")
if insights["top_tags"]:
lines.extend(["## 高频标签 TOP 15", ""])
for tag, count in insights["top_tags"].items():
lines.append(f"- {tag} ({count}次)")
lines.append("")
if insights["top5_feeds"]:
lines.extend(["## TOP 5 高赞笔记", ""])
lines.append("| 排名 | 标题 | 作者 | 赞数 |")
lines.append("|------|------|------|------|")
for i, f in enumerate(insights["top5_feeds"], 1):
lines.append(f"| {i} | {f['title']} | {f['author']} | {f['likes']} |")
lines.append("")
lines.extend([
"---",
"",
"## 创作建议",
"",
"基于以上数据,创作时优先参考:",
])
if insights["title_types"]:
top_title = list(insights["title_types"].keys())[0]
lines.append(f"- 标题采用 **{top_title}** 效果最佳")
if insights["image_styles"]:
top_style = list(insights["image_styles"].keys())[0]
lines.append(f"- 配图优先用 **{top_style}** 风格")
if insights["top_elements"]:
top3_elems = list(insights["top_elements"].keys())[:3]
lines.append(f"- 必备要素:{', '.join(top3_elems)}")
lines.append("")
with open(SUMMARY_FILE, "w", encoding="utf-8") as f:
f.write("\n".join(lines))
print(f"\n摘要已保存到: {SUMMARY_FILE}", file=sys.stderr)
def delete(feed_id: str):
"""删除一条笔记"""
db = _load_db()
before = len(db["feeds"])
db["feeds"] = [f for f in db["feeds"] if f["feed_id"] != feed_id]
after = len(db["feeds"])
_save_db(db)
print(json.dumps({"success": True, "deleted": before - after, "remaining": after}))
def stats():
"""输出数据库基本统计"""
db = _load_db()
feeds = db["feeds"]
domains = Counter(f.get("domain", "未分类") for f in feeds)
print(json.dumps({
"total": len(feeds),
"domains": dict(domains.most_common()),
"db_path": str(DB_FILE),
"db_size": DB_FILE.stat().st_size if DB_FILE.exists() else 0,
}, ensure_ascii=False, indent=2))
def main():
if len(sys.argv) < 2:
print("用法: feed_database.py <command> [args]")
print("命令:")
print(" add '<json>' 添加/更新笔记")
print(" list [--domain X] [--min-likes N] [--limit N]")
print(" get <feed_id> 获取单条详情")
print(" delete <feed_id> 删除")
print(" analyze 生成分析摘要")
print(" stats 数据库统计")
sys.exit(0)
cmd = sys.argv[1]
if cmd == "add":
if len(sys.argv) < 3:
# 从 stdin 读取
feed_json = sys.stdin.read()
else:
feed_json = sys.argv[2]
add(feed_json)
elif cmd == "list":
kwargs = {}
args = sys.argv[2:]
i = 0
while i < len(args):
if args[i] == "--domain" and i + 1 < len(args):
kwargs["domain"] = args[i + 1]
i += 2
elif args[i] == "--min-likes" and i + 1 < len(args):
kwargs["min_likes"] = int(args[i + 1])
i += 2
elif args[i] == "--limit" and i + 1 < len(args):
kwargs["limit"] = int(args[i + 1])
i += 2
else:
i += 1
list_feeds(**kwargs)
elif cmd == "get":
get(sys.argv[2])
elif cmd == "delete":
delete(sys.argv[2])
elif cmd == "analyze":
analyze()
elif cmd == "stats":
stats()
else:
print(f"未知命令: {cmd}")
sys.exit(1)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |