Dan Romero
@dwr.eth
Farcaster dry powder * Problem What is the aggregate dollar value of crypto assets ("dry powder") held in all protocol-linked addresses? * Output 1. A GitHub repo / gist with a reproducible script (assume API key) to generate the figure; include instructions for how to run, avoid rate limits, etc. 2. Tiebreaker: a good Dune dashboard with the figure * Criteria Bias for first correct submission, but will give everyone at least 24 hours to submit before making a decision. * Amount 500 USDC @bountycaster
23 replies
15 recasts
75 reactions
yesyes
@yesyes
Since no one has answered yet - https://github.com/realguy33/FC-balances/blob/main/balances.ipynb This is very basic. it outputs value of the total balance(in USD) in all addresses divided by total number of unique FIDs, Also, it was optimized a bit with the help of LLMs. I am currently cross checking to ensure the output is correct(seems to be correct so far) Let me know if you were looking for something like this . If so, then i will streamline and optimize it more as well as include the solana data.
3 replies
0 recast
1 reaction
yesyes
@yesyes
I preferred to do it using the dune's API and not Dune's SQL because some prices tables in dune SQL are highly unreliable. This API fetches more accurate prices data. It also uses pagination so technically SOME coins that are on the next page of an address's token list will not be included but they are liquidity spam coins doesn't make much of a difference.(The api lists highly liquid coins before the lower liquidity ones). i can completely exclude the low liquidity coins too which will speed up the query even more. Let me know if you were thinking in this direction. Will invest more time in it if it seems worthwhile. Let me know if this is the direction
0 reply
0 recast
2 reactions
yesyes
@yesyes
oh i noticed that i accidentally had the native tokens filter on. let me remove that real quick lol(the code only lists ETH and matic for now)
1 reply
0 recast
0 reaction
yesyes
@yesyes
Update - it now analyzes all tokens. not just native tokens. removed low liquidity ones for more accurate data. Unfortunately, the LLM optimization is inaccurate. Will see tomorrow why. For now, i have included the much slower but accurate method 2. Again, very basic stuff, don't want to spend too much time on this. Hoping that someone else may have submitted some solution by tomorrow.
1 reply
0 recast
1 reaction
