Cold Email AI Reply Agent

What it does

Pulls unread prospect replies from your cold email platform, classifies each into clear categories (interested, not interested, out of office, etc.) and applies one exclusive funnel tag per reply. Drafts a short, human-sounding reply, respecting daily caps and avoiding autoresponders.

When to use it

• When you get high reply volume and can't keep up classifying by hand.
• When you want consistent funnel tagging without error-prone manual rules.
• When you need fast replies but still reviewed before they go out.

Steps

1. The assistant fetches unread replies from your cold email platform.
2. It classifies each reply into a predefined category.
3. It applies the matching funnel tag without overlapping others.
4. It drafts a short, human-sounding reply.
5. It respects daily send caps and flags autoresponders to avoid replying.
6. You approve each reply before it sends; nothing goes out without your sign-off.

The skill

---
name: cold-email-reply-agent
description: Classifies prospect replies into categories, applies funnel tags, and drafts short human-sounding replies with safety limits.
---

# Cold Email AI Reply Agent

## 1. What it does

- Pulls unread prospect replies from your cold email platform, classifies each into clear categories (interested, not interested, out of office, etc.) and applies one exclusive funnel tag per reply. Drafts a short, human-sounding reply, respecting daily caps and avoiding autoresponders.

## 2. When to use it

- When you get high reply volume and can't keep up classifying by hand.
- When you want consistent funnel tagging without error-prone manual rules.
- When you need fast replies but still reviewed before they go out.

## 3. Step by step

- The assistant fetches unread replies from your cold email platform.
- It classifies each reply into a predefined category.
- It applies the matching funnel tag without overlapping others.
- It drafts a short, human-sounding reply.
- It respects daily send caps and flags autoresponders to avoid replying.
- You approve each reply before it sends; nothing goes out without your sign-off.

## 4. Safety rules

- Safety rules: everything ships as a draft first; nothing is sent, published, or paid for without your explicit approval; no data, numbers, or results are ever invented; no credentials or secrets are typed into chat.

## 5. Best practices

- Always review the full draft before approving any send, publish, or payment.
- If real information (data, numbers, results) is missing, ask for it to be flagged as pending instead of invented.
- Keep credentials and keys out of chat; set them in environment variables or your secrets manager.

## 6. GoHighLevel integration

- Where relevant, results can be saved as contacts, tags, tasks, or notes inside your GoHighLevel sub-account.
- No automation inside GHL (workflow, campaign, trigger) activates without your review first.

For Claude Code, save it as ~/.claude/skills/cold-email-reply-agent/SKILL.md. In Cursor, ChatGPT or another assistant, paste it as project instructions.