agentleFS
Kuan-AI-comKuan-AI-com/agent-skills/skills/support-reply/SKILL.md

support-reply

Skill0 starsChanged 6 days agoMITView on GitHub

Draft a reply to a customer support message, ticket, review or chat. Use when the user pastes a customer message and asks how to respond, "answer this ticket", "reply to this complaint", or wants a support macro or template.

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the policy, FAQ, docs or macros available in the project or conversation (refunds, shipping, SLAs, warranty).

    line 18
  2. Distinctive:

    If information is missing (order number, email, screenshot, device), ask for all of it in one list.

    line 28

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Customer support reply

1. Triage

  • What does the customer want (refund, fix, information, to be heard)? List every issue if there are several.
  • Type: bug, how-to question, billing, shipping/delivery, account access, feature request, complaint.
  • Mood: calm, frustrated, angry, threatening to leave.

2. Escalate instead of answering alone when the message involves

legal threats or lawyers, chargebacks or fraud, safety or injury, a security vulnerability or account compromise, data deletion/access requests (GDPR, CCPA), press or public social media posts, or a refund/credit above what the policy lets you grant. Then draft a short holding reply (received, who will follow up, when if known) and an internal note saying why it needs a human.

3. Check the source of truth

Read the policy, FAQ, docs or macros available in the project or conversation (refunds, shipping, SLAs, warranty). If none are available, write the reply with [policy: ...] placeholders. Never:

  • promise anything the policy does not allow
  • invent timelines ("within 24 hours") without a stated SLA
  • state account facts you cannot see (order status, tracking, plan) -> use [order status]
  • say an action is done ("I've issued a refund") unless the user told you it is done. Write "I've requested..." or [confirm refund issued].

4. Draft

  • First sentence names their specific problem, not "we apologize for any inconvenience". Apologise once if we are at fault; do not admit fault for things not yet confirmed.
  • The answer or fix. Steps as a numbered list, one action per step, using the exact names of buttons and menus from the docs.
  • What happens next and who acts: "You'll get an email from us when...".
  • If information is missing (order number, email, screenshot, device), ask for all of it in one list.
  • Under 120 words unless steps require more. Customer's language. No internal jargon or ticket codes. Never blame the customer.
  • Sign with [Agent name].

5. Internal note (separate, only if useful)

Bug to report (steps, frequency), doc gap, policy exception requested, or churn risk.

Output

  1. Triage line: type, mood, escalate yes/no.
  2. The reply.
  3. Placeholders to fill.
  4. Internal note, if any.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.