email-brain-classifier

Classifies gbrain email pages into categories, detects risk signals, and routes to department profiles.

1|Updated Jun 21, 2026
One-click install
npx skills add https://github.com/tapway/shogun-os --skill email-brain-classifier-tapway
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: email-brain-classifier
Source: https://github.com/tapway/shogun-os/tree/main/skills/gbrain/email-brain-classifier
Command: npx skills add https://github.com/tapway/shogun-os --skill email-brain-classifier-tapway

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? After emails are ingested into gbrain, they sit unprocessed with no category, no owner, and no risk visibility. This Skill automates the triage of those email pages so project updates, deal activity, support issues, and HR correspondence reach the right department profile, while high-risk messages trigger immediate Slack alerts. ## Core Features & Use Cases - Five-category classification: Sorts emails into project, deal, support, HR, or other using signal keywords, then routes each to the correct brain path and owning profile. - Risk and sentiment detection: Scans every email for HIGH risk signals (legal threats, outages, escalations, payment disputes) and MEDIUM risk signals (scope creep, vendor delays), sending Slack alerts to management for HIGH risk items. - Brain routing and cross-linking: Resolves matching projects or deals, adds timeline entries, tags emails as classified, and creates bidirectional links so no content is orphaned. - Use Case: A batch of 20 newly imported emails includes an angry customer escalation and a purchase order. The Skill alerts #management about the escalation, files the PO under the matching active project, and reports a processing summary. ## Quick Start Ask the agent to run the email brain classifier on all unprocessed email pages in gbrain and report the category breakdown and any risk alerts.

Frequently Asked Questions about email-brain-classifier

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automatically classify emails in gbrain?

Run this classifier as a post-collection processor after emails are imported. It queries gbrain for email pages without the classified tag, reads each page, matches content against a five-category keyword taxonomy, then tags and routes each email to the correct brain path.

How does email risk detection and alerting work?

Every email is scanned for risk signals before routing. HIGH risk signals like legal threats, outages, or escalations trigger an immediate Slack alert to the management channel, while MEDIUM risk items are logged to a risk register page in gbrain.

Can this classifier run without MCP gbrain tools?

Yes. In subagent or terminal-only contexts where MCP tools are unavailable, it falls back to gbrain CLI commands such as gbrain search, gbrain tag, and gbrain timeline-add. New pages can be written directly to the brain directory followed by gbrain sync.

When should I not use the email brain classifier?

Do not use it for initial email collection, which belongs to the ingest pipeline or Gmail collector scripts, or for meeting notes, which have a dedicated meeting classifier. It also does not replace manual reading and replying through the Gmail integration.

Why does the classifier create duplicate project or deal pages?

Duplicates happen when the resolution step is skipped. The workflow requires searching gbrain for an existing project or deal by customer name and keywords before creating any new page, then cross-linking the email to the matched page.