Do you feel like you've curated a good Crypto Twitter? Put your money where your likes are! With HummingScroller, you sit and zone-out on Twitter, liking crypto posts you agree with, as you already do. HummScroller watches your Twitter account and buys the tokens (present in $cashtags
) in the Tweets that you hit "like" on!
After 24 hours, HummScroller then sells that token back into stablecoins, and pulls your new recent likes to build a new portfolio. New portfolios are built every day based on the changing sentiment towards coins that you decide by liking them.
Install it Via Telegram: TBD
Telegram Bot Command: /hummscroller @my_handle
-- call the bot and enter your Twitter handle.
This registers a paper "mock" trading account in a Hummingbot instance.
Now just keep zoning out on Crypto Twitter! It's scanning tweets that you like.
Currently only working on paper trading with a Binance connector through Hummingbot.
Your most recent 100 Tweets will be scanned for any cashtags, which will then build a dataset about what you've liked recently.
Set Twitter API Keys:
export TWITTER_HANDLE=elonmusk
-- you can monitor other people's Twitter likes too!
export TWITTER_API_BEARER=AAAAA...
-- need a Twitter Developer account to use V2 API.
Twitter Script: humscroller.py
-- produces a file recent_tweet_likes.json
Hummingbot Script: humminggram.py
-- Performs daily market order trades based on the data.
Got to this point where it was buying BTC-USDT, but wasn't updating the paper balance on Hummingbot
From Discord: you can see the trade status
, but there's a bug not reporting the paper balance.
TODO: https://hummingbot.org/operation/export/ ... Export the trades to a file and then make that data available from a Telegram bot, so you can see your daily tweet-like-trade performances.
Telegram Bot: TBD -- commands like /humscroller_init @handle
, /humscroller_stats
, humscroller_leaderboard
Hosting & Deployment: TBD
- A Telegram Group "Investment DAO" Game where you compete with members to have the more profitable Crypto Twitter.
- Countertrading. Not your twitter handle, because you're smart, but someone who you think is an idiot. Everything they like, you sell.
- Use real money, not paper trading.
- Scan not just likes, but retweets also.
- AI sentiment classification to break tweets into the bullish and bearish parts.
- Get audio from
Spaces
you were in, how long you stayed (enjoyed it) and correlate that to tokens.