# Mail Triage

> >-

- **Type:** Skill
- **Install:** `agentstack add skill-aashari-ai-agent-skills-mail-triage`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [aashari](https://agentstack.voostack.com/s/aashari)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [aashari](https://github.com/aashari)
- **Source:** https://github.com/aashari/ai-agent-skills/tree/main/skills/apple-mail/mail-triage

## Install

```sh
agentstack add skill-aashari-ai-agent-skills-mail-triage
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Mail Triage — Smart Priority Overview

Triaging: **$ARGUMENTS** (default: last 48 hours)

## Step 1: Pull all candidate emails
```bash
DB="$HOME/Library/Mail/V10/MailData/Envelope Index"
SINCE=$(($(date +%s) - 172800))  # 48 hours; adjust per $ARGUMENTS

sqlite3 "$DB" "
SELECT datetime(m.date_received,'unixepoch','localtime') as dt,
       s.subject, a.address as sender, a.comment as name,
       mb.url as mailbox, m.ROWID,
       m.read, m.flagged, m.is_urgent, m.size, m.conversation_id,
       m.automated_conversation, m.unsubscribe_type
FROM messages m
JOIN subjects  s  ON m.subject = s.ROWID
JOIN addresses a  ON m.sender  = a.ROWID
JOIN mailboxes mb ON m.mailbox = mb.ROWID
WHERE m.date_received >= ${SINCE}
  AND m.deleted = 0
  AND mb.url NOT LIKE '%Spam%' AND mb.url NOT LIKE '%Trash%'
  AND mb.url NOT LIKE '%Junk%' AND mb.url NOT LIKE '%Sent%'
ORDER BY m.date_received DESC;" 2>/dev/null
```

## Step 2: Score each email for priority

**High priority signals (+points):**
- is_urgent = 1: +10
- flagged = 1: +8
- read = 0: +3
- Subject contains: urgent, ASAP, action required, critical, deadline, review, approve: +5 each
- automated_conversation = 0 AND unsubscribe_type = 0 (likely real person): +4
- From work/EWS account: +2
- Large message (>50KB): +1 (likely real content)
- Subject contains reply indicators (RE:, FW:): +2

**Low priority signals (−points):**
- automated_conversation = 2 (bulk automated — newsletters, monitoring, CI/CD): −8
- unsubscribe_type > 0 (has List-Unsubscribe header — mailing list): −5
- Subject starts with PROBLEM:, OK:, ALARM:, CRIT: −8 (monitoring)
- Already read: −5

**Compute a score per email, then bucket:**
- Score ≥ 10: Critical
- Score 5–9: Important
- Score 1–4: Normal
- Score ≤ 0: Noise

## Step 3: Read bodies for top-priority items
For Critical and Important emails (score ≥ 5):
```bash
python3 ~/.claude/skills/_mail-shared/parser.py   ...
```

Extract key info: what is being asked, any deadlines, any amounts.

## Step 4: Check for conversation context
For unread replies, check if it's a continuation of a thread you started:
```bash
sqlite3 "$DB" "
SELECT COUNT(*) FROM messages m
JOIN mailboxes mb ON m.mailbox=mb.ROWID
WHERE m.conversation_id=
  AND mb.url LIKE '%Sent%';" 2>/dev/null
```
If you sent something in this thread → higher priority (someone replied to you).

## Output Format
**Critical — Act Now:**
[list with 1-line context from body]

**Important — Handle Today:**
[list with subject and sender]

**Normal — Read When Convenient:**
[count + brief list]

**Noise (filtered):** X emails
[breakdown: Y monitoring alerts, Z newsletters, W CI/CD]

Total: X emails in window. Estimated action time: [estimate based on count].
Offer to read any specific email or start replying.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [aashari](https://github.com/aashari)
- **Source:** [aashari/ai-agent-skills](https://github.com/aashari/ai-agent-skills)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-aashari-ai-agent-skills-mail-triage
- Seller: https://agentstack.voostack.com/s/aashari
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.