Skip to content

A nostr proxy relay that prevents spam (based on your criteria and judged using AI) from being written to your relay. Project created on the 2023 SatsHack.

Notifications You must be signed in to change notification settings

johnnyasantoss/orangecheckr

Repository files navigation

OrangeProxy

SatsHack (Hackaton from SatsConf 2023) project based on a proxy to help relay operators maintain a community that is spam free and aligned with their private policy of what is allowed to be posted (eg. can't that about shitcoins in this relay, etc.).

Leveraging nip42 to prevent attackers and onboard users on the relay (without leaving the client) and LLMs to parse the privacy policy and to judge if the content adheres to the policy.

The idea was inspired by Michael Saylor's OrangeCheck 1 2

Sub projects

  • proxy: intercepts messages to ensure that only users that are authenticated (NIP42) with a committed collateral can stay connected on the relay, and helps the bot to onboard new users
  • bot: onboard new users on the relay using DMs
  • llm: parses the content policy (to be written by the relay operator) and judges if the content adheres to it or not. L402 TBD
  • dashboard: TBD page where the relay operators can see what was reported and filter/ban users from the relay

Using it

The code was written in ~24h by sleep deprived developers. BE AWARE

Having your relay online (implementation agnostic), setup the proxy in front of it and setup your .env file with the required configs/secrets. Create a nostr key to be used as the onboarding bot. Host the llm and setup .env accordingly

TBD

Footnotes

  1. https://twitter.com/saylor/status/1418590438781554689

  2. https://twitter.com/saylor/status/1508244604726001675

About

A nostr proxy relay that prevents spam (based on your criteria and judged using AI) from being written to your relay. Project created on the 2023 SatsHack.

Topics

Resources

Stars

Watchers

Forks