评论洞察：一次调用，直接拿到结论

参数

返回什么

字段与平台无关：帖子类 items[]（url、文案、作者、时间、互动数据、封面），达人类 items[]（handle、昵称、简介、粉丝数、认证），洞察类直接给 sentiment / questions / top_phrases / top_liked / summary。

在 MCP 客户端里怎么用

不需要记参数——把端点加进客户端后，直接用自然语言描述需求，模型会自己挑工具和参数。写代码时再用上面的 REST 形式。

复制即可运行

curl -s https://lucydomi.com/mcp/v1/get_comment_insights \
  -H "Authorization: Bearer $BUSYBEE_API_KEY" \
  -H "Content-Type: application/json" \
  -X POST \
  -d '{"platform": "tiktok", "post_id": "7690199760430566669", "sample": 20}'

import os, json, urllib.request

KEY = os.environ["BUSYBEE_API_KEY"]
req = urllib.request.Request(
    "https://lucydomi.com/mcp/v1/get_comment_insights",
    data=json.dumps({"platform": "tiktok", "post_id": "7690199760430566669", "sample": 20}).encode(),
    headers={"Authorization": f"Bearer {KEY}", "Content-Type": "application/json"},
    method="POST",
)
data = json.load(urllib.request.urlopen(req, timeout=60))
for it in data["items"]:
    print(it.get("url"), it.get("text", "")[:60])

真实返回样例

下面这段是线上真实返回（已截断到 1 条，去掉了计费字段）。

{
  "total": 20,
  "sentiment": {
    "positive": 2,
    "negative": 0,
    "neutral": 18,
    "positive_hits": 3,
    "negative_hits": 0
  },
  "questions": {
    "count": 9,
    "ratio": 0.45
  },
  "engagement": {
    "max_likes": 1103,
    "avg_likes": 78.4
  },
  "top_phrases": [
    {
      "phrase": "spanish",
      "count": 5
    },
    {
      "phrase": "coffee",
      "count": 3
    },
    {
      "phrase": "watery",
      "count": 3
    },
    {
      "phrase": "ice",
      "count": 3
    },
    {
      "phrase": "honey",
      "count": 2
    },
    {
      "phrase": "latte",
      "count": 2
    },
    {
      "phrase": "spain",
      "count": 2
    },
    {
      "phrase": "sooo",
      "count": 2
    },
    {
      "phrase": "good",
      "count": 2
    }
  ],
  "top_liked": [
    {
      "text": "Make it three",
      "likes": 1103,
      "author": "Samantha"
    },
    {
      "text": "how do u guys prevent the coffee from being watery? when i put a lot of ice it gets watery fast, or am i just drinking it too slowly 🥲",
      "likes": 145,
      "author": "lex"
    },
    {
      "text": "The honey? the salt? the cinnamon??? Disastrous.",
      "likes": 135,
      "author": "bubibat"
    },
    {
      "text": "Where’s the name “Spanish latte” coming from? Because this is not a normal coffee in Spain. It would be a café bombón",
      "likes": 52,
      "author": "angelagvarela"
    },
    {
      "text": "What makes it Spanish?",
      "likes": 31,
      "author": "Arturo Hlebov"
    }
  ],
  "summary": "共分析 20 条评论；其中 9 条（45%）是提问，说明购买意图明确；情绪正面为主（正面 2 条 / 负面 0 条）；高频词：spanish、coffee、watery、ice、honey；单条最高点赞 1103",
  "post": {
    "platform": "tiktok",
    "id": "7690199760430566669",
    "text": null,
    "stats": {}
  },
  "ok": true,
  "platform": "tiktok",
  "tool": "get_comment_insights",
  "sampled": 20
}

常见问题

相关内容