From agi-super-team
Triggers the Telegram scraper agent for testing or unscheduled channel scanning. Supports dry-run, category filtering, and notification tests.
How this skill is triggered — by the user, by Claude, or both
Slash command
/agi-super-team:telegram-scraper-runThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
> Runs the Telegram Scraper Agent manually for testing or unscheduled scanning.
Runs the Telegram Scraper Agent manually for testing or unscheduled scanning.
Optional:
--dry-run - test run without notifications--category <name> - scan only one category (competitors/industry/advertising)--no-messages - skip reading messages (faster)--notify-test - notification test onlycd $AGENTS_PATH/telegram-scraper
python3 telegram_scraper_agent.py
# Test without notifications
python3 telegram_scraper_agent.py --dry-run
# Single category
python3 telegram_scraper_agent.py --category competitors --dry-run
# Without reading messages (faster)
python3 telegram_scraper_agent.py --no-messages --dry-run
python3 telegram_scraper_agent.py --notify-test
python3 test_telegram_scraper.py
Agent outputs:
Data is saved to:
$PROJECT_ROOT/data/telegram_scraper/
├── YYYY-MM-DD/ # Dated results
│ ├── competitors_channels.json
│ ├── competitors_ad_contacts.csv
│ ├── industry_channels.json
│ ├── advertising_channels.json
│ └── messages/
└── latest/ # Symlinks to most recent
# Latest results
ls -l $PROJECT_ROOT/data/telegram_scraper/latest/
# Top 5 channels (competitors)
cat $PROJECT_ROOT/data/telegram_scraper/latest/competitors_channels.json | jq '.[0:5]'
# Ad contacts
cat $PROJECT_ROOT/data/telegram_scraper/latest/competitors_ad_contacts.csv
# Agent log
cat $PROJECT_ROOT/data/telegram_scraper/agent_log.json | jq '.[-5:]'
Edit config:
code $PROJECT_ROOT/data/telegram_scraper_config.json
Config structure:
{
"categories": {
"competitors": {
"keywords": ["annotation", "data labeling", "cvat"],
"exclude": ["spam", "crypto"],
"scan_posts": 10
}
},
"min_subscribers": 100,
"min_score": 10,
"notification_threshold": 30
}
Agent runs automatically twice daily (9:00, 18:00).
# Check status
launchctl list | grep telegram-scraper
# Load schedule
launchctl load ~/Library/LaunchAgents/com.yourcompany.telegram-scraper.plist
# Unload schedule
launchctl unload ~/Library/LaunchAgents/com.yourcompany.telegram-scraper.plist
# View logs
tail -f $GOOGLE_TOOLS_PATH/logs/telegram_scraper.log
tail -f $GOOGLE_TOOLS_PATH/logs/telegram_scraper.err
If Telegram session is invalid:
# Refresh session
cd $TG_TOOLS_PATH
python3 -m tg_utils.auth
cd $TG_TOOLS_PATH && python3 -m tg_utils.auth--dry-run for debugscan_posts in configIf the agent is not working:
cd $TG_TOOLS_PATH/tools
# Find channels with ad contacts
python3 tg_scrape.py ads --keywords "annotation,labeling" --posts 10
# List channels
python3 tg_scrape.py channels --keywords "ai,ml" --output channels.csv
# Read messages
python3 tg_scrape.py messages "Channel Name" --days 7 --limit 50
After scraping:
add-lead skilltelegram-send skilltelegram-session - update Telegram sessionadd-lead - add found contacts to CRMtelegram-send - message ad contactsdaily-briefing - include findings in morning briefingnpx claudepluginhub aaaaqwq/agi-super-team --plugin agi-super-teamSearch Telegram channels by keywords, read messages, and extract ad contacts from channel descriptions and posts. Useful for channel discovery and ad analysis.
Search Telegram channels and pull audience/engagement stats (subscribers, reach, ER/ERR) via TGStat API. Useful for finding channels to advertise in and researching competitor audiences.
Parses public Telegram channels via t.me/s/ web previews to fetch posts, metrics, analytics, digests, top posts, and schedules. Enables monitoring and competitive analysis using pure bash scripts.