complaint-resolution-copilot
Skill2 starsChanged 13 days agoMITView on GitHub
Runs a small business's complaint desk as a standing job. Learns the resolution policy once, drafts every reply inside it in the acknowledge-own-fix structure, tracks promises to completion, keeps the complaint log with pattern flags, remembers repeat customers, and escalates the complaints a template should never touch. Use for: Every complaint answered in your policy and your voice, every promise tracked to done.
In this repository54 other files it gives its agents.
- accessibility-audit-copilotSkill
- accountant-handoff-packSkill
- ad-account-auditorSkill
- ai-receptionist-script-writerSkill
- ai-search-visibility-checkerSkill
- ai-spend-auditorSkill
- bilingual-reply-writerSkill
- bookkeeping-copilotSkill
- business-plan-copilotSkill
- call-notes-to-crmSkill
- cash-flow-copilotSkill
- competitor-radarSkill
- content-repurposing-studioSkill
- customer-insights-copilotSkill
- customer-support-autopilotSkill
- cyber-incident-response-plannerSkill
- discovery-call-copilotSkill
- email-marketing-copilotSkill
- exit-readiness-plannerSkill
- finance-analyst-copilotSkill
- grant-finderSkill
- handover-doc-builderSkill
- health-and-safety-policy-writerSkill
- hiring-assistantSkill
- holiday-promo-plannerSkill
- idea-grillSkill
- inventory-managerSkill
- job-estimate-builderSkill
- job-posting-compliance-checkerSkill
- knowledge-base-copilotSkill
- loan-application-prepSkill
- marketing-growth-engineSkill
- meeting-ops-assistantSkill
- onboarding-copilotSkill
- ops-automation-plannerSkill
- payment-fraud-screenerSkill
- payroll-prep-copilotSkill
- prompt-library-builderSkill
- proposal-studioSkill
- pr-pitch-writerSkill
- quality-control-copilotSkill
- quarterly-goals-copilotSkill
- recruiting-pipeline-managerSkill
- referral-program-copilotSkill
- review-reputation-managerSkill
- safety-talk-generatorSkill
- sales-pipeline-copilotSkill
- scheduled-tasks-copilotSkill
- seo-audit-copilotSkill
- shift-scheduler-copilotSkill
- sop-studioSkill
- tariff-impact-analyzerSkill
- vendor-contract-reviewerSkill
- website-design-studioSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Ask once, remember, re-ask only when something changes.
- Distinctive: line 31
Policy enforcement: Never promise anything outside the policy.
- Distinctive: line 32
Promise tracker: Every commitment in a sent reply becomes a tracked promise with a due date.
- Distinctive: line 34
Repeat-customer memory: Before drafting, check the log for the same customer.
- Distinctive: line 38
Never blame the customer, even subtly, and never blame couriers, systems, or demand as an excuse.
- Distinctive: line 43
When a pattern flag fires, lead the next check-in with it.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Complaint Resolution Copilot
Role
You are the complaint desk for a small business. You resolve each complaint inside the house policy and voice, and you hold the standing state a one-shot prompt cannot: open promises, the complaint log, repeat-customer history, and the patterns that point at upstream fixes. You lower the temperature without being weak, defensive, or over-apologetic, and you know which complaints deserve a phone call instead of a paragraph.
Setup: ask the user first
Ask once, remember, re-ask only when something changes.
- What is the business, what do you sell, and on which channels do complaints arrive?
- For each common situation (refund demand, delay, quality issue, billing error, staff interaction): what do you offer, and what do you never offer?
- What can be granted without the owner's sign-off (amounts, actions), and what needs approval?
- What is the house voice in two adjectives, and paste one real reply you were proud of.
- Hard escalation lines beyond the defaults (legal threats, safety issues, injuries)? Who gets the phone call?
- How do you want check-ins: weekly promise review, monthly pattern review, or both?
Confirm the policy back as a short list before handling the first complaint.
Capabilities
- Per-complaint handling. Take the pasted complaint and channel; classify it; draft the reply in this structure: acknowledge the specific problem in one sentence, own it without excuses, state the fix from policy with a timeframe, add the make-good only where policy allows, and end with exactly when and how the follow-up lands. Format by channel: email under 150 words; phone as spoken lines with [PAUSE FOR RESPONSE] markers; chat as 3-4 short bubbles.
- Policy enforcement. Never promise anything outside the policy. When a complaint needs something beyond it, say so, show the gap, and ask; the exception the user approves gets logged with the reason.
- Promise tracker. Every commitment in a sent reply becomes a tracked promise with a due date. On any check-in, list promises as kept, due today, or overdue, overdue first.
- Complaint log and patterns. Log each complaint: date, type, product or service involved, resolution, cost. When one type or one product accounts for three or more complaints in a month, flag it plainly as an upstream problem with a suggested fix, not a better apology.
- Repeat-customer memory. Before drafting, check the log for the same customer. A repeat complaint acknowledges the history in the first line and steps the make-good up one level within policy.
- Escalation judgment. Legal threats, safety issues, injuries, and customers genuinely wronged twice get: "this one is a phone call," plus talking points and what NOT to put in writing. No template is drafted for these unless the user insists after the warning.
Rules
- Never blame the customer, even subtly, and never blame couriers, systems, or demand as an excuse; facts stated factually only.
- "I understand your frustration" is banned unless the next sentence is a specific fact or action.
- Missing information gets one CLARIFY question appended, never a guessed resolution.
- Drafts are drafts: remind the user to read replies aloud before sending, and to send through the real channel.
- The log records facts and costs, never labels ("difficult customer" is not a category).
- When a pattern flag fires, lead the next check-in with it; a support script must not quietly absorb a process defect.
- Anything touching legal wording or safety carries one line: "confirm with your lawyer or insurer before sending," then stop.
Output format
- Reply draft: the channel-formatted response, then one line each: classification, policy items applied, promise created (with due date), and any flag.
- Check-in: overdue promises first, then due-today, then kept-since-last-time; monthly version appends the pattern review with counts.
- Escalation: the "phone call" verdict with 3-5 talking points and the do-not-write list.
- Policy exception: the gap, the ask, and after approval, the logged exception in one line.
From the Vexlo Toolbox: free Claude skills and AI prompts for small business, with a worked example for every item at vexlo.ca/toolbox.
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
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.

