An ongoing responsibility - the collection of Sybil and non-Sybil lists

Recently the FDD from Gitcoin, our founding sponsor and inspiration, has ceased operations effective the end of April 2023. Gitcoin has decided to instead distribute data analysis and related responsibilities amongst existing work groups within Gitcoin or to the OpenData Community.

One of the responsibilities that the OpenData Community has agreed to assume is the collection, curation and protection of reference data sets for likely Sybils and non-Sybils.

These lists are commonly referred to as Loki - for the list of likely Sybils - and Thor - for the list of likely non-Sybils.

Via this post, I would like to summarize our initial approach and intent regarding these important resources and solicit feedback from the broader community.

  1. Approach:

Access controls: We intend to secure these lists so that anyone with a legitimate need can gain access to them, including active ODC members whether Regen Rangers or ODC Jedi, and all active hackathon participants. Any disclosure or abuse of these lists will be considered a violation of the code of conduct of the OpenData Community and of all hackathons it supports. We are currently storing these lists on the Ocean Protocol - accessible to a whitelist of known wallets - at:

Additions: We also intend to add to these lists, typically by adding additional related data sets derived from subsequent analyses. For example, in the Regen Rangers hackathon in April of 2023 we are providing to contestants access to other lists of likely Sybils in addition to the original Loki list, including all those wallets squelched as likely Sybils in the January 2023 Alpha round from Gitcoin.

Appeals: We anticipate that in rare cases wallets will request to be removed from the lists. We are currently staffed solely by volunteers. As such, our responses to any appeals will be on a best-efforts basis. Nonetheless, it is our intention to be responsive and transparent to any appeals.

  1. Intent

It is our intent to do what we can to tilt conditions slightly back in the favor of those resisting fraud and Sybil attacks across web3. As such we want our community to be able to judge the efficacy of Legos and other algorithms with the aid of these reference data sets.

Summary drawing:

The above schematic summarizes our current approach.


In conclusion, we wish to manage lists of Sybils and non-Sybils as a public good that can be used by the broader web3 ecosystem. We welcome additional donations of data and offers of assistance. Your feedback on our approach would also be appreciated.