Some notes
- Fluxbot is a new Telegram trading bot for the Solana ecosystem. It is a self custodial wallet.
- The private key can be imported in another wallet. We successfully imported it to phantom wallet.
- It uses Jupiter for DCA and other dex for swapping (Raydium, fluxbeam, etc…)
- We can’t modfiy slippage for every transaction.
- We can’t see a readable transaction history, we have to go to https://solscan.io/
FLUXB token usage:
The native token is not used for transaction fees, so there is no real demand for the token. However, holders of at least 1000 FLUXB token get “reflections”. In short, they get rewarded with a part of the fees generated by the platform. The more there are accounts above 1000, the less your balance becomes compared to the total number of FluxB pool so the less rewards you get. This will incentivize people to hold more tokens,
Reflections depend on the success of the platform, so marketing and adoption are a must. What distinguishes Fluxbot is the copy trader and sniper feature. However as more platforms start offering the same features for the Solana ecosystem, the fees will have to be lowered to stay competitive which will decrease the reflections as well.
What sets apart fluxb is the AI chatbot part. If the chatbot becomes more interactive and personal, users will become more attached to using it, so this is one path towards success.
We went on to test the AI assistant.
In the first example, we asked the assistant to convert 1$ worth of solana to fluxb token:
It replied insufficient balance although we have 1$ worth of solana in the balance. Probably it though we meant 1 SOLANA, so we rephrased it with the amount in solana:
This time it understood and prepared the right transaction. The next step is to test wether the AI bot remembers our discussion, so let’s ask him to buy FLYXB using PYTH, but let’s do it a bit vaguely to see the level of interaction:
We see that the bot assumed we wanted to buy pyth instead of fluxb, not buy pyth instead of fluxb. Our statement was vague and confusing and we expected the bot to ask which option it should choose. However we were a bit more specific:
We can see that the bot is following closely the conversion and was able to correct the transaction.
Overall, the assistant is satisfactory for a first version and some more improvements should be on the way.
We tried to take it a step further and ask the bot to modify the slippage on the fly:
But it still used the predefined slippage from the settings.