# Twitter Search

> Search Twitter/X for crypto sentiment, influencer opinions, and community discussions. Activate when researching tokens, projects, or market trends.

- Skill: `juntoku9/twitter-search` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add juntoku9/twitter-search`
- Raw SKILL.md: https://api.skillmd.com/api/skills/juntoku9/twitter-search/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: juntoku9 (https://skillmd.com/u/juntoku9)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/juntoku9/twitter-search

---


# Twitter/X Search for Crypto Research

## Overview

Search Twitter/X for real-time crypto sentiment and discussions. Essential for:
- Token sentiment analysis ($BTC, $ETH, $SOL)
- Influencer opinions and alpha
- Community reactions to news
- Trending narratives

## Usage

```bash
python3 .claude/skills/twitter-search/scripts/search_twitter.py "$SOL" --count 20
```

## Quality Filters (NEW!)

Use these flags to get higher quality tweets:

```bash
# Minimum engagement filters
--min-likes 50        # Only tweets with 50+ likes (high quality signal)
--min-retweets 10     # Only tweets with 10+ retweets
--min-replies 5       # Only tweets with 5+ replies

# Content filters
--lang en             # Only English tweets
--no-retweets         # Exclude retweets (original content only)
--verified            # Only from verified accounts
--has-media           # Only tweets with images/videos
--has-links           # Only tweets with links

# Time filters
--since 2025-01-01    # Tweets after this date
--until 2025-01-31    # Tweets before this date
--within 7d           # Last 7 days (also: 24h, 30m)
```

## Crypto Search Examples

```bash
# High quality token discussion (50+ likes, no retweets, English)
python3 .claude/skills/twitter-search/scripts/search_twitter.py "\$ARB" --min-likes 50 --no-retweets --lang en --count 20

# Verified accounts talking about a project
python3 .claude/skills/twitter-search/scripts/search_twitter.py "Arbitrum" --verified --min-likes 20 --count 15

# Recent high engagement tweets (last 7 days)
python3 .claude/skills/twitter-search/scripts/search_twitter.py "\$SOL" --within 7d --min-likes 100 --type Top --count 15

# Filter out spam/noise with engagement
python3 .claude/skills/twitter-search/scripts/search_twitter.py "\$ETH" --min-likes 10 --no-retweets --count 20

# Media posts only (charts, infographics)
python3 .claude/skills/twitter-search/scripts/search_twitter.py "OpenMind tokenomics" --has-media --min-likes 20 --count 15

# Specific user's tweets
python3 .claude/skills/twitter-search/scripts/search_twitter.py --user VitalikButerin --count 10

# Combined: high quality discussion in last 24h
python3 .claude/skills/twitter-search/scripts/search_twitter.py "OpenMind airdrop" --min-likes 30 --within 24h --lang en --no-retweets --count 20
```

## Output Format

```
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
@username (Display Name) ✓
Followers: 125K | Jan 15, 2025 14:30

Tweet content here...

❤️ 1.2K  🔁 567  💬 89  👁️ 45.7K
https://twitter.com/username/status/123456789
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
```

## What to Capture

- **Sentiment**: Is overall mood bullish, bearish, or neutral?
- **Influencer takes**: What are CT thought leaders saying?
- **Community concerns**: What worries are being discussed?
- **Excitement**: What's driving positive sentiment?
- **Narratives**: What stories/themes are trending?

## Recommended Search Strategy for Researchers

**For General Sentiment** (balanced view):
```bash
python3 ... "PROJECT_NAME" --min-likes 10 --no-retweets --lang en --count 20
```

**For Quality Alpha** (influencers, high engagement):
```bash
python3 ... "\$TOKEN" --min-likes 50 --verified --type Top --count 15
```

**For Recent News** (last 24-48h):
```bash
python3 ... "PROJECT_NAME announcement" --within 48h --min-likes 20 --count 15
```

**For Visual Content** (charts, infographics):
```bash
python3 ... "PROJECT_NAME metrics" --has-media --min-likes 30 --count 10
```

## Search Tips

- Always use quality filters (--min-likes at minimum) to avoid spam
- Use `--no-retweets` for original opinions/analysis
- Use `--lang en` to get English content
- For tokens, use `$SYMBOL` format (e.g., `$SOL`, `$ETH`)
- Combine `--verified` with `--min-likes` for high-signal content
- Use `--type Top` for most engaging tweets
- Use `--within 7d` for recent, relevant discussions