Drop Hyacinthe into Claude and get an email security engineer who gets you to DMARC reject without breaking payroll, and finds the mailbox rule hiding the fraud. Hyacinthe defends the channel most breaches still start in: email authentication done properly with SPF flattening and the ten-lookup limit, DKIM key rotation and selector hygiene, and DMARC taken from none to reject without cutting off a legitimate vendor, plus BIMI; gateway and cloud-native tuning on Microsoft Defender for Office 365, Proofpoint, Mimecast and Abnormal; phishing, spear-phishing and business email compromise including invoice redirect and payroll diversion; account takeover detection in Microsoft 365 and Google Workspace covering illicit OAuth consent grants, forwarding rules and the inbox rules that hide the fraud; attachment and URL detonation and its bypasses; collaboration security across Teams, Slack, SharePoint, OneDrive and Google Drive with external sharing sprawl and guest access; the phish-report triage queue; quarantine and release governance; and the tenant hardening baseline that removes whole classes of attack. What you get →SPF, DKIM and DMARC to reject without breaking senders →Gateway tuning for phishing, BEC and payroll diversion →Microsoft 365 and Workspace account-takeover detection →External sharing, guest access and tenant hardening baseline 📄 hyacinthe-email-collaboration-security-engineer.skill Under 2 min install Works with Claude, ChatGPT & any AI chat How to install Download the .skill package, open Claude, paste SKILL.md into your Project Instructions or system prompt, describe your requirement, and Hyacinthe builds the answer. Includes a full worked example so you see exactly what you get.
