Preface

Upon receiving a customer service ticket, the first step is not to draft a reply, but to determine whether it falls under inquiry, complaint, expediting, or returns/exchanges. You must assess the customer’s current emotional state (D1 or D4), identify the deadline for the first response, recognize prohibited language, and determine escalation scenarios. These judgments are scattered across script libraries, SLA tables, and quality inspection rules; relying on human memory to synthesize them often leads to omissions. The Skill introduced below solidifies these steps into a pipeline: it inputs the original ticket text, customer sentiment, order details, and policy guidelines, and outputs a structured customer service reply accompanied by an SLA review ledger. This is suitable for teams building their own customer service middle office or using large language models for ticket replies.

What is this

Customer Service Reply Generator Pro, published by Kunlun Growth, is sourced from enterprise, categorized under business-ops, with the current version 1.1.0, and coordinates @kunlungrowth/kunlun-cs-reply-pro. It solves the problem of transforming scattered rules—such as ticket classification, sentiment grading, first response thresholds, banned word verification, escalation determination, and review logging—into a workflow that can be completed in a single call.

Core Functions

According to the official description, the capabilities are organized into a four-stage process:

  1. First, determine the ticket classification and D1-D4 sentiment level.
  2. Next, apply the first response threshold, perform banned word verification, and execute the escalation decision tree.
  3. Output a four-section customer service reply.
  4. Attach an SLA review ledger.

Trigger words cover: customer service reply, complaint, expediting, return/exchange, first response, SLA, negative review, and scripts. The tag requires_api_key=false indicates that no additional API Key configuration is needed for usage.

Installation and Activation

Official documentation does not provide installation commands, directory structures, or configuration file fields. The page only provides the coordinate @kunlungrowth/kunlun-cs-reply-pro and the SkillHub page address. According to SkillHub conventions, you can search and install it via the coordinate on the SkillHub page; please refer to the real-time instructions on the page for specific commands, as this article does not speculate on them. Regarding potential differences between Cursor, Codex CLI, and Claude Code, the documentation does not mention them, so this cannot be verified and is therefore omitted.

Typical Usage

The input and output relationship described in the official summary is as follows, which can serve as a starting point for reproduction:

Input:

Original ticket text
Customer sentiment
Order information
Policy guidelines

The output is organized into four sections: first, it provides the ticket classification and D1-D4 sentiment level; second, it applies the first response threshold, banned word verification, and the escalation decision tree; finally, it provides the customer service reply body and SLA review ledger entries. The trigger method relies on keyword matching; hitting any of the keywords—customer service reply, complaint, expediting, return/exchange, first response, SLA, negative review, or scripts—should activate the Skill.

Applicable Scenarios and Notes

This is suitable for customer service teams with clear SLA standards, strict scripting red lines, and established escalation rules, especially in scenarios where reply quality and response timeliness are assessed together.

Before use, you need to confirm a few things:

  1. Policy guidelines must be maintained independently: The Skill is only responsible for applying them and does not provide the source of the policies.
  2. Escalation decision tree configuration: Rules for the escalation decision tree need to be configured according to actual business operations, and the determination results depend on the completeness of these rules.
  3. Banned word coverage: The coverage of banned word verification depends on the configured vocabulary list and requires regular updates.
  4. Adoption Rate: The page shows 0 installations, indicating there is no public data on usage volume yet. It is recommended to conduct a small-scale trial before a full rollout.

Conclusion

The value of this Skill lies in transforming ticket replies from “manual judgment” to “rule-driven, four-section output” while conveniently leaving an SLA review ledger. For teams that already have a customer service middle office and script library, the integration cost mainly involves completing the policy guidelines and decision tree configuration.

Official Address: https://www.skillhub.cn/skills/kunlungrowth/kunlun-cs-reply-pro
API Address: https://api.skillhub.cn/kunlungrowth/kunlun-cs-reply-pro