CipherOwl

By CipherOwl

Overview

CipherOwl brings its blockchain address screening into ChatGPT for compliance, risk, and investigations teams. Ask about a cryptocurrency address and the plugin can: - Screen it: report whether CipherOwl found risk for the address under your organization's risk configuration. - Explain it: list the reasons behind a finding, such as sanctions, hacks, theft, or scams, with the direction, number of hops, and value of the exposure. Supported networks include Ethereum and other EVM chains, Bitcoin, TRON, and Solana. Which networks you can screen depends on your organization's CipherOwl plan. You sign in with your CipherOwl account and choose your organization. Results use that organization's data access, permissions, and risk configuration. The plugin is read-only. It cannot send, freeze, or report funds, and it does not create alerts or cases. CipherOwl's results are risk signals that support your own compliance decisions. They are not legal advice, and the plugin does not identify private individuals.

Tools

reason_risk

ChatGPT

Use after or instead of screen when per-category risk evidence is needed. Returns read-only categories, directions, hops, values, percentages, and labels, plus a links entry with the CipherOwl page for the address on the chains that page supports — include that link when the response carries one, and never construct one yourself; returns NO_RESULTS when no risk is found. On a restricted plan the response also carries plan: tell the user in this answer what plan.explanation says the plan limits, so a limited result is never presented as complete.

screen

ChatGPT

Use for a read-only sanctions and risk verdict on one blockchain address. Returns the resolved risk configuration, verdict, risk level, and explanations, plus a links entry with the CipherOwl page for the address on the chains that page supports — include that link when the response carries one, and never construct one yourself. Safe to retry; alert tracking is always disabled. On a restricted plan the response also carries plan: tell the user in this answer what plan.explanation says the plan limits, so a limited result is never presented as complete.