Sales & Customer Support Skills Kit — Installation Guide

This pack includes 10 ready-to-use Claude Skills for outreach, sales conversations, and customer support:

1. cold-outreach-email-writer — short, personalized first-touch sales emails
2. sales-proposal-summary-writer — turn deal details into a client-facing proposal summary
3. sales-objection-handler — structure responses to specific objections
4. discovery-call-question-planner — structured discovery-call question sets
5. follow-up-sequence-planner — multi-touch follow-up sequences after a meeting or no response
6. renewal-outreach-email-writer — renewal and upsell outreach for existing customers
7. win-loss-summary-writer — structure a win/loss deal summary for internal review
8. customer-support-reply-writer — empathetic, solution-focused support replies
9. support-ticket-triage-categorizer — categorize and prioritize a batch of support tickets
10. customer-onboarding-email-sequence-planner — plan a post-sale onboarding email sequence

What is a Claude Skill?

A Claude Skill is an instructions file (SKILL.md) that teaches Claude how to perform a specific task following your format and rules. When you ask Claude something that matches a skill's description, it automatically applies those instructions.

Installing — Claude Code (on your computer / terminal)

A) User-level (works in every project)

1. Create the skills folder if it doesn't exist yet:

   mkdir -p ~/.claude/skills
   

2. Copy each skill's folder into ~/.claude/skills/, so the final path looks like:

   ~/.claude/skills/cold-outreach-email-writer/SKILL.md
   ~/.claude/skills/sales-proposal-summary-writer/SKILL.md
   ~/.claude/skills/sales-objection-handler/SKILL.md
   ~/.claude/skills/discovery-call-question-planner/SKILL.md
   ~/.claude/skills/follow-up-sequence-planner/SKILL.md
   ~/.claude/skills/renewal-outreach-email-writer/SKILL.md
   ~/.claude/skills/win-loss-summary-writer/SKILL.md
   ~/.claude/skills/customer-support-reply-writer/SKILL.md
   ~/.claude/skills/support-ticket-triage-categorizer/SKILL.md
   ~/.claude/skills/customer-onboarding-email-sequence-planner/SKILL.md
   

3. Start (or restart) Claude Code — the skills will be available automatically in any project.

B) Single-project level

Copy the same folders into <your-project>/.claude/skills/ instead, if you only want them in one project.

Using with claude.ai / Claude Desktop

1. Open Settings.
2. Find the Capabilities or Skills section.
3. Upload each skill's SKILL.md file (or its whole folder).
4. The skill triggers automatically when you ask something matching its description.

Using with ChatGPT

ChatGPT doesn't have a "Skills" system, but each SKILL.md file works just as well as ChatGPT instructions — the part below the --- frontmatter is plain instruction text.

As a Custom GPT (ChatGPT Plus/Team — stays saved for reuse):

1. Go to Explore GPTs → Create → Configure.
2. Give it a name (e.g., "Cold Outreach Writer") and a short description.
3. Open the skill's SKILL.md file, copy everything below the --- frontmatter block, and paste it into the Instructions field.
4. Save. Reopen this GPT any time you need that skill.

Without Custom GPT access (works on any ChatGPT plan): Paste the same instructions (below the frontmatter) at the start of a new chat, then ask your question. You'll need to re-paste it each time you start a new chat.

How to check it's working

After installing, ask Claude something that clearly matches one skill, for example:

"Write a cold outreach email to a Head of Marketing at an ecommerce brand, offering our email automation tool."

If the skill triggered correctly, the answer will follow the cold-outreach-email-writer structure (specific opener, short body, a low-friction ask, under 120 words).

Common issues

• The skill doesn't trigger — check that the SKILL.md file is inside a folder, not dropped directly into .claude/skills/.
• Changes aren't showing up — restart your Claude Code session after editing a file.
• Output feels generic — the more specific context you give (actual prospect details, what actually happened on the call), the more specific and useful the output will be.