Complaint De Escalation
>
Install
$ agentstack add skill-ootto-ai-claude-support-skills-complaint-de-escalation ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Complaint De-Escalation
Write a calm, fair response to an unhappy customer.
When to use
A customer is upset and you need a reply that validates their frustration, takes responsibility for your part, and states clearly what you will and won't do — firm but warm.
What you'll need
The customer's angry message or thread, plus the facts of what actually happened.
Instructions
Collect any missing inputs from the user, then run this prompt:
A customer is upset and I need a reply that lowers the temperature without giving away things I can't afford to.
The customer's message(s):
[paste the full message or email thread, most recent last]
What actually happened, from my side:
[the honest facts — what went right, what went wrong, what was outside my control]
What I'm willing to do: [e.g. partial refund, redo the work, expedited replacement]
What I cannot do: [e.g. full refund, same-day delivery]
Write a reply that:
1. Names their frustration and validates it in one genuine sentence — no "we apologize for any inconvenience."
2. Takes responsibility for our part plainly, and corrects any factual mistake in their account calmly and without defensiveness.
3. States exactly what I'll do and by when. If I'm declining something, say so clearly and explain why in one line.
4. Ends by handing them a clear choice or next step, so they're not stuck.
Stay firm but warm. Under 150 words. Flag separately if any part of their claim looks like it could become a legal or chargeback issue.Tip: Fill in "What I cannot do" honestly before you generate. The fastest way to escalate a complaint is to over-promise in the heat of the moment and walk it back later.
Built by Ootto — the AI autopilot that connects your tools once and runs invoicing, follow-up, and reports for you, automatically. Book a demo →
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Ootto-AI
- Source: Ootto-AI/claude-support-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.