nexscope-ai/eCommerce-Skills / brand-monitoring
brand-monitoring
VerifiedBrand monitoring tool for tracking mentions across social media platforms. Monitor Reddit, Google News, YouTube, and DuckDuckGo for brand mentions. Includes sentiment analysis, trend tracking, crisis detection, and competitor comparison. No API key required for basic monitoring.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.9 KB |
| scripts/monitor.py | script | 22.4 KB |
| scripts/scrapers.py | script | 7.9 KB |
scripts/monitor.py
View on GitHub#!/usr/bin/env python3
"""
Brand Monitoring - Core Engine
Brand Monitoring - Core Engine
Features:
- BrandMention Monitoring
- SentimentAnalyze (positive/negative/ in )
- TrendTrack
- Abnormal/Crisis Detection
- CompetitorVolumeComparison
- Keyword Extraction
- ReportGenerate
Datasource (Lite):
- Reddit (PublicSearch)
- Google ( new news/Blog)
- YouTube (Review)
- ManualInput
Version: 1.0.0
"""
import json
import re
from dataclasses import dataclass, field
from typing import List, Dict, Optional, Tuple
from enum import Enum
from datetime import datetime, timedelta
from collections import Counter
import sys
class Sentiment(Enum):
"""SentimentCategoryType"""
POSITIVE = "positive"
NEGATIVE = "negative"
NEUTRAL = "neutral"
class Platform(Enum):
"""PlatformCategoryType"""
REDDIT = "reddit"
TWITTER = "twitter"
YOUTUBE = "youtube"
AMAZON = "amazon"
GOOGLE = "google"
TIKTOK = "tiktok"
FACEBOOK = "facebook"
FORUM = "forum"
NEWS = "news"
OTHER = "other"
class AlertLevel(Enum):
"""AlertLevelother"""
INFO = "info"
WARNING = "warning"
CRITICAL = "critical"
# ============================================================
# SentimentKeywordsLibrary
# ============================================================
POSITIVE_KEYWORDS = {
"en": [
"love", "great", "amazing", "awesome", "excellent", "perfect", "best",
"fantastic", "wonderful", "recommend", "happy", "satisfied", "quality",
"worth", "impressed", "reliable", "favorite", "brilliant", "superb"
],
"zh": [
"Like", " very good ", "great", "Excellent", "Perfect", " most good ", "Recommended", "Satisfied", "Quality good ",
"Value ", "Impressive", "Canrely", " most love", "Excellent"
]
}
NEGATIVE_KEYWORDS = {
"en": [
"hate", "terrible", "awful", "worst", "bad", "horrible", "disappointed",
"waste", "scam", "fake", "poor", "broken", "useless", "regret", "avoid",
"never", "refund", "complaint", "problem", "issue", "sucks"
],
"zh": [
"Hate", "Terrible", "difference", " most difference", " bad ", "Disappointed", "Waste", "Scammer", "Counterfeit",
"Poor", " bad ", "Useless", " after regret", "Avoid", "Refund", "Complaint", "Issue"
]
}
CRISIS_KEYWORDS = {
"en": [
"lawsuit", "recall", "investigation", "fraud", "scandal", "danger",
"safety", "warning", "banned", "illegal", "death", "injury", "toxic"
],
"zh": [
"Lawsuit", "Recall", "Investigation", "Fraud", "Scandal", "Danger", "safe all ", "Warning",
"Prohibited", "Illegal", "Death", "Harm", " has toxic"
]
}
# ============================================================
# Data Structures
# ============================================================
@dataclass
class Mention:
"""Brandmention and """
content: str
platform: Platform
url: str = ""
author: str = ""
date: Optional[datetime] = None
engagement: int = 0 # Interactivequantity (likes, comments, shares)
sentiment: Sentiment = Sentiment.NEUTRAL
sentiment_score: float = 0.0
keywords: List[str] = field(default_factory=list)
is_influential: bool = False
@dataclass
class BrandMetrics:
"""BrandMetrics"""
total_mentions: int = 0
positive_count: int = 0
negative_count: int = 0
neutral_count: int = 0
avg_sentiment_score: float = 0.0
total_engagement: int = 0
top_keywords: List[Tuple[str, int]] = field(default_factory=list)
platform_distribution: Dict[str, int] = field(default_factory=dict)
daily_trend: Dict[str, int] = field(default_factory=dict)
@dataclass
class Alert:
"""Alert"""
level: AlertLevel
title: str
title_zh: str
description: str
description_zh: str
trigger: str
timestamp: datetime = field(default_factory=datetime.now)
@dataclass
class CompetitorComparison:
"""CompetitorComparison"""
brand: str
metrics: BrandMetrics
share_of_voice: float = 0.0
@dataclass
class MonitoringReport:
"""MonitoringReport"""
brand_name: str
period_start: datetime
period_end: datetime
metrics: BrandMetrics
mentions: List[Mention]
alerts: List[Alert]
competitors: List[CompetitorComparison]
top_positive: List[Mention]
top_negative: List[Mention]
summary: str
summary_zh: str
# ============================================================
# SentimentAnalyze
# ============================================================
def analyze_sentiment(text: str, lang: str = "en") -> Tuple[Sentiment, float]:
"""AnalyzeTextSentiment"""
text_lower = text.lower()
positive_score = 0
negative_score = 0
# KeywordsMatch
for keyword in POSITIVE_KEYWORDS.get(lang, []) + POSITIVE_KEYWORDS.get("en", []):
if keyword.lower() in text_lower:
positive_score += 1
for keyword in NEGATIVE_KEYWORDS.get(lang, []) + NEGATIVE_KEYWORDS.get("en", []):
if keyword.lower() in text_lower:
negative_score += 1
# CalculateScore (-1 to +1)
total = positive_score + negative_score
if total == 0:
return Sentiment.NEUTRAL, 0.0
score = (positive_score - negative_score) / max(total, 1)
if score > 0.2:
sentiment = Sentiment.POSITIVE
elif score < -0.2:
sentiment = Sentiment.NEGATIVE
else:
sentiment = Sentiment.NEUTRAL
return sentiment, round(score, 2)
def extract_keywords(text: str, top_n: int = 5) -> List[str]:
"""ExtractKeywords"""
# SimpleWord frequencyStatistics
words = re.findall(r'\b[a-zA-Z]{4,}\b', text.lower())
# FilterStop words
stopwords = {"this", "that", "with", "have", "from", "they", "been", "were", "will", "would", "could", "should", "about", "their", "there", "which", "when", "what", "your", "just", "some", "more", "very", "really"}
words = [w for w in words if w not in stopwords]
counter = Counter(words)
return [word for word, count in counter.most_common(top_n)]
def detect_crisis(text: str) -> bool:
"""DetectionCrisisKeywords"""
text_lower = text.lower()
for keyword in CRISIS_KEYWORDS.get("en", []):
if keyword in text_lower:
return True
return False
# ============================================================
# DataCollect (Lite - Simulate/Demo)
# ============================================================
def search_reddit(brand_name: str, limit: int = 20) -> List[Mention]:
"""Search Reddit (Lite: SimulateData)"""
# In real implementationοΌ this in will Use Reddit API or Crawler
# Lite versionUse pushshift.io or CategorySeemingly free service
print(f"[Reddit] Searching for: {brand_name}")
return []
def search_google_news(brand_name: str, limit: int = 20) -> List[Mention]:
"""Search Google News (Lite: SimulateData)"""
print(f"[Google News] Searching for: {brand_name}")
return []
def get_demo_mentions(brand_name: str) -> List[Mention]:
"""GetDemoData"""
now = datetime.now()
return [
Mention(
content=f"Just got my {brand_name} product and I absolutely love it! Best purchase this year.",
platform=Platform.REDDIT,
url="https://reddit.com/r/product/xxx",
author="happy_customer",
date=now - timedelta(hours=2),
engagement=150,
),
Mention(
content=f"Has anyone tried {brand_name}? Looking for honest reviews before buying.",
platform=Platform.REDDIT,
url="https://reddit.com/r/product/yyy",
author="curious_buyer",
date=now - timedelta(hours=5),
engagement=45,
),
Mention(
content=f"Disappointed with {brand_name} quality. The product broke after 2 weeks. Waste of money!",
platform=Platform.TWITTER,
url="https://twitter.com/user/status/xxx",
author="angry_buyer",
date=now - timedelta(hours=8),
engagement=320,
),
Mention(
content=f"Comparing {brand_name} vs competitors - honest review video coming soon!",
platform=Platform.YOUTUBE,
url="https://youtube.com/watch?v=xxx",
author="tech_reviewer",
date=now - timedelta(days=1),
engagement=5000,
),
Mention(
content=f"{brand_name} announces new product line with improved features",
platform=Platform.NEWS,
url="https://news.example.com/article",
author="TechNews",
date=now - timedelta(days=1),
engagement=200,
),
Mention(
content=f"I've been using {brand_name} for 6 months now. Highly recommend for anyone looking for quality.",
platform=Platform.REDDIT,
url="https://reddit.com/r/reviews/xxx",
author="long_time_user",
date=now - timedelta(days=2),
engagement=89,
),
Mention(
content=f"Warning: {brand_name} customer service is terrible. Been waiting 3 weeks for refund!",
platform=Platform.TWITTER,
url="https://twitter.com/user/status/yyy",
author="frustrated_customer",
date=now - timedelta(days=2),
engagement=180,
),
Mention(
content=f"Just saw {brand_name} at the store. The packaging looks premium!",
platform=Platform.TIKTOK,
url="https://tiktok.com/@user/video/xxx",
author="shopping_vlog",
date=now - timedelta(days=3),
engagement=2500,
),
]
# ============================================================
# CoreAnalyze
# ============================================================
def analyze_mentions(mentions: List[Mention]) -> Tuple[List[Mention], BrandMetrics]:
"""Analyzemention and Data"""
# SentimentAnalyze
for mention in mentions:
sentiment, score = analyze_sentiment(mention.content)
mention.sentiment = sentiment
mention.sentiment_score = score
mention.keywords = extract_keywords(mention.content)
mention.is_influential = mention.engagement > 1000
# CalculateMetrics
metrics = BrandMetrics()
metrics.total_mentions = len(mentions)
metrics.positive_count = len([m for m in mentions if m.sentiment == Sentiment.POSITIVE])
metrics.negative_count = len([m for m in mentions if m.sentiment == Sentiment.NEGATIVE])
metrics.neutral_count = len([m for m in mentions if m.sentiment == Sentiment.NEUTRAL])
if mentions:
metrics.avg_sentiment_score = round(sum(m.sentiment_score for m in mentions) / len(mentions), 2)
metrics.total_engagement = sum(m.engagement for m in mentions)
# PlatformDistribution
for mention in mentions:
platform = mention.platform.value
metrics.platform_distribution[platform] = metrics.platform_distribution.get(platform, 0) + 1
# KeywordsStatistics
all_keywords = []
for mention in mentions:
all_keywords.extend(mention.keywords)
metrics.top_keywords = Counter(all_keywords).most_common(10)
return mentions, metrics
def generate_alerts(mentions: List[Mention], metrics: BrandMetrics) -> List[Alert]:
"""GenerateAlert"""
alerts = []
# NegativeSentimentProportionAlert
if metrics.total_mentions > 0:
negative_ratio = metrics.negative_count / metrics.total_mentions
if negative_ratio > 0.4:
alerts.append(Alert(
level=AlertLevel.CRITICAL,
title="High Negative Sentiment",
title_zh="NegativeSentiment High",
description=f"Negative mentions at {negative_ratio*100:.1f}% - above 40% threshold",
description_zh=f"Negativemention and Proportion {negative_ratio*100:.1f}% - super 40% Threshold",
trigger=f"negative_ratio > 0.4",
))
elif negative_ratio > 0.3:
alerts.append(Alert(
level=AlertLevel.WARNING,
title="Elevated Negative Sentiment",
title_zh="NegativeEmotion riseHigh",
description=f"Negative mentions at {negative_ratio*100:.1f}% - above 30% threshold",
description_zh=f"Negativemention and Proportion {negative_ratio*100:.1f}% - super 30% Threshold",
trigger=f"negative_ratio > 0.3",
))
# HighImpactpowerNegativemention and
influential_negative = [m for m in mentions if m.is_influential and m.sentiment == Sentiment.NEGATIVE]
if influential_negative:
alerts.append(Alert(
level=AlertLevel.CRITICAL,
title="Influential Negative Mention",
title_zh="HighImpactpowerNegativemention and ",
description=f"Found {len(influential_negative)} negative mention(s) with high engagement",
description_zh=f"Found {len(influential_negative)} itemHighInteractivequantityNegativemention and ",
trigger="engagement > 1000 AND sentiment = negative",
))
# CrisisKeywordsDetection
crisis_mentions = [m for m in mentions if detect_crisis(m.content)]
if crisis_mentions:
alerts.append(Alert(
level=AlertLevel.CRITICAL,
title="Potential Crisis Detected",
title_zh="potential in Crisis Detection",
description=f"Found {len(crisis_mentions)} mention(s) with crisis keywords",
description_zh=f"Found {len(crisis_mentions)} Contains crisisKeywordsmention and ",
trigger="crisis_keywords detected",
))
return alerts
def monitor_brand(
brand_name: str,
competitors: List[str] = None,
mentions: List[Mention] = None,
) -> MonitoringReport:
"""Brand MonitoringMainFunction"""
now = datetime.now()
period_start = now - timedelta(days=7)
# GetData
if mentions is None:
mentions = get_demo_mentions(brand_name)
# Analyze
mentions, metrics = analyze_mentions(mentions)
# GenerateAlert
alerts = generate_alerts(mentions, metrics)
# CompetitorComparison
competitor_data = []
if competitors:
total_mentions = metrics.total_mentions
for comp in competitors:
comp_mentions = get_demo_mentions(comp)[:3] # Demo: few quantityData
_, comp_metrics = analyze_mentions(comp_mentions)
sov = comp_metrics.total_mentions / (total_mentions + comp_metrics.total_mentions) if total_mentions > 0 else 0
competitor_data.append(CompetitorComparison(
brand=comp,
metrics=comp_metrics,
share_of_voice=round(sov, 2),
))
# Top mention and
sorted_positive = sorted([m for m in mentions if m.sentiment == Sentiment.POSITIVE], key=lambda x: -x.engagement)
sorted_negative = sorted([m for m in mentions if m.sentiment == Sentiment.NEGATIVE], key=lambda x: -x.engagement)
# Summary
pos_pct = metrics.positive_count / metrics.total_mentions * 100 if metrics.total_mentions > 0 else 0
neg_pct = metrics.negative_count / metrics.total_mentions * 100 if metrics.total_mentions > 0 else 0
summary = f"π {metrics.total_mentions} mentions | π {pos_pct:.0f}% positive | π {neg_pct:.0f}% negative | β οΈ {len(alerts)} alerts"
summary_zh = f"π {metrics.total_mentions} Item mention and | π {pos_pct:.0f}% Positive | π {neg_pct:.0f}% Negative | β οΈ {len(alerts)} itemAlert"
return MonitoringReport(
brand_name=brand_name,
period_start=period_start,
period_end=now,
metrics=metrics,
mentions=mentions,
alerts=alerts,
competitors=competitor_data,
top_positive=sorted_positive[:3],
top_negative=sorted_negative[:3],
summary=summary,
summary_zh=summary_zh,
)
# ============================================================
# OutputFormat
# ============================================================
def format_report(report: MonitoringReport, lang: str = "en") -> str:
"""FormatReport"""
m = report.metrics
if lang == "zh":
lines = [
"π‘ **Brand MonitoringReport**",
"",
f"**Brand**: {report.brand_name}",
f"**MonitoringweeksPeriod**: {report.period_start.strftime('%Y-%m-%d')} ~ {report.period_end.strftime('%Y-%m-%d')}",
"",
"ββββββββββββββββββββββββββββββββββββ",
"",
"## π OverallMetrics",
"",
f"| Metrics | Value |",
f"|------|------|",
f"| Total mention and | {m.total_mentions} |",
f"| Positivemention and | {m.positive_count} ({m.positive_count/m.total_mentions*100:.0f}%) |" if m.total_mentions > 0 else f"| Positivemention and | 0 |",
f"| Negativemention and | {m.negative_count} ({m.negative_count/m.total_mentions*100:.0f}%) |" if m.total_mentions > 0 else f"| Negativemention and | 0 |",
f"| Neutralmention and | {m.neutral_count} |",
f"| AverageSentiment score | {m.avg_sentiment_score} |",
f"| Total engagement | {m.total_engagement:,} |",
"",
]
# Alert
if report.alerts:
lines.extend([
"ββββββββββββββββββββββββββββββββββββ",
"",
"## β οΈ Alert",
"",
])
for alert in report.alerts:
level_icon = "π¨" if alert.level == AlertLevel.CRITICAL else "β οΈ" if alert.level == AlertLevel.WARNING else "βΉοΈ"
lines.append(f"**{level_icon} {alert.title_zh}**")
lines.append(f" {alert.description_zh}")
lines.append("")
# PlatformDistribution
if m.platform_distribution:
lines.extend([
"ββββββββββββββββββββββββββββββββββββ",
"",
"## π± PlatformDistribution",
"",
])
for platform, count in sorted(m.platform_distribution.items(), key=lambda x: -x[1]):
pct = count / m.total_mentions * 100 if m.total_mentions > 0 else 0
lines.append(f"- **{platform}**: {count} ({pct:.0f}%)")
lines.append("")
# PopularKeywords
if m.top_keywords:
lines.extend([
"ββββββββββββββββββββββββββββββββββββ",
"",
"## π PopularKeywords",
"",
", ".join([f"`{word}` ({count})" for word, count in m.top_keywords[:8]]),
"",
])
# Top Positivemention and
if report.top_positive:
lines.extend([
"ββββββββββββββββββββββββββββββββββββ",
"",
"## π PopularPositivemention and ",
"",
])
for i, mention in enumerate(report.top_positive[:3], 1):
lines.append(f"**{i}. [{mention.platform.value}]** Interactive: {mention.engagement:,}")
lines.append(f" \"{mention.content[:80]}...\"")
lines.append("")
# Top Negativemention and
if report.top_negative:
lines.extend([
"ββββββββββββββββββββββββββββββββββββ",
"",
"## π need AttentionNegativemention and ",
"",
])
for i, mention in enumerate(report.top_negative[:3], 1):
lines.append(f"**{i}. [{mention.platform.value}]** Interactive: {mention.engagement:,}")
lines.append(f" \"{mention.content[:80]}...\"")
lines.append("")
lines.extend([
"ββββββββββββββββββββββββββββββββββββ",
"",
report.summary_zh,
])
else:
# English version
lines = [
"π‘ **Brand Monitoring Report**",
"",
f"**Brand**: {report.brand_name}",
f"**Period**: {report.period_start.strftime('%Y-%m-%d')} ~ {report.period_end.strftime('%Y-%m-%d')}",
"",
"ββββββββββββββββββββββββββββββββββββ",
"",
"## π Overview Metrics",
"",
f"| Metric | Value |",
f"|--------|-------|",
f"| Total Mentions | {m.total_mentions} |",
f"| Positive | {m.positive_count} ({m.positive_count/m.total_mentions*100:.0f}%) |" if m.total_mentions > 0 else "| Positive | 0 |",
f"| Negative | {m.negative_count} ({m.negative_count/m.total_mentions*100:.0f}%) |" if m.total_mentions > 0 else "| Negative | 0 |",
f"| Neutral | {m.neutral_count} |",
f"| Avg Sentiment | {m.avg_sentiment_score} |",
f"| Total Engagement | {m.total_engagement:,} |",
"",
]
if report.alerts:
lines.extend(["## β οΈ Alerts", ""])
for alert in report.alerts:
level_icon = "π¨" if alert.level == AlertLevel.CRITICAL else "β οΈ"
lines.append(f"**{level_icon} {alert.title}**: {alert.description}")
lines.append("")
if m.platform_distribution:
lines.extend(["## π± Platform Distribution", ""])
for platform, count in sorted(m.platform_distribution.items(), key=lambda x: -x[1]):
lines.append(f"- **{platform}**: {count}")
lines.append("")
lines.extend(["", report.summary])
return "\n".join(lines)
# ============================================================
# CLI
# ============================================================
def main():
lang = "zh" if "--zh" in sys.argv else "en"
brand = sys.argv[1] if len(sys.argv) > 1 and not sys.argv[1].startswith("--") else "TechBrand"
report = monitor_brand(brand, competitors=["CompetitorA", "CompetitorB"])
print(format_report(report, lang))
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirements | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |