Review Reply Coach

v1.0.0

Generates professional, tone-appropriate reply templates to guide users in responding to customer reviews with thanks, acknowledgment, apology, or de-escalat...

0· 40·0 current·0 all-time
byhaidong@harrylabsj

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for harrylabsj/review-reply-coach.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Review Reply Coach" (harrylabsj/review-reply-coach) from ClawHub.
Skill page: https://clawhub.ai/harrylabsj/review-reply-coach
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install review-reply-coach

ClawHub CLI

Package manager switcher

npx clawhub@latest install review-reply-coach
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name, description, prompts, and skill.json all describe the same capability (generating reply templates and internal policy documents). There are no required binaries, environment variables, or config paths that are unrelated to the stated purpose.
Instruction Scope
SKILL.md instructs the agent only to classify review tone and generate reply templates, escalation paths, and multi-language outputs. It explicitly forbids fabricating reviews, manipulating reviews, or sharing PII in public replies. The instructions do not request access to unrelated files, environment variables, or external endpoints.
Install Mechanism
No install spec and no code files — the skill is instruction-only, which minimizes surface area. README suggests an installer command for the platform but the skill bundle contains no external download or install instructions that would write code to disk.
Credentials
No environment variables, credentials, or config paths are required. The skill's claimed functionality does not need secrets or system access, so the lack of requested credentials is proportionate.
Persistence & Privilege
Flags indicate default behavior (always:false, model invocation allowed). The skill does not request permanent presence or modify other skills' configurations. Autonomous invocation is allowed by default but is not combined with any broad credential or install activity.
Assessment
This skill appears coherent and low-risk: it only contains prompts and safety rules for composing responses to existing reviews and requests no credentials or installs. Before installing, remember: (1) never paste customers' private data into prompts; (2) if you plan to automate posting replies, keep posting credentials separate and review that workflow for permissions and auditing; and (3) the skill.json claims no network/code execution, but that is declarative — verify the platform enforces those constraints if you require strict isolation.

Like a lobster shell, security has layers — review code before you run it.

latestvk97bc72dmnmdt9pzgrzvbgpnrh85mxet
40downloads
0stars
1versions
Updated 17h ago
v1.0.0
MIT-0

Compliant Review Reply Coach

Purpose

This skill generates professional, tone-calibrated responses to customer reviews across e-commerce and app store platforms. Unlike other content-generator skills that create new content, this is a service skill — it coaches the user on how to reply appropriately to existing reviews. It covers three response modes: positive (thank + reinforce brand value), neutral (acknowledge + commit to improvement), and negative (apologize + resolve + take conversation offline). "Coach" emphasizes guidance, templates, and best-practice education — not automated mass-replying. This is the only skill in the pack that does NOT generate new marketing content.

Triggers

  • "回复评论"
  • "review response"
  • "reply to customer review"
  • "差评回复"
  • "app review reply"
  • "商家回复"
  • "review reply coach"
  • "客户评论回复"
  • "好评怎么回"
  • "中评怎么回"

Workflow

  1. Receive the review text, star rating (1–5), and platform context from the user.
  2. Classify the review tone into one of four categories: Positive (4–5★, praising), Neutral (3★, mixed feelings), Negative (1–2★, disappointed), Angry (1★, emotional/aggressive).
  3. Select the appropriate response framework:
    • Positive: Thank specifically → Reinforce what they liked → Invite future engagement → Subtle upsell (optional, disclosed)
    • Neutral: Thank for feedback → Acknowledge the mixed experience → Commit to specific improvement → Offer follow-up
    • Negative: Apologize sincerely (without admitting fault if unverified) → Address specific complaint → Offer concrete resolution → Take conversation offline (email/phone)
    • Angry: De-escalate first → Acknowledge emotion → Offer resolution path → Escalate if safety/legal issue
  4. Apply the CRITICAL safety gate: verify the output is a REPLY to an existing review, never a fabricated review.
  5. For serious complaints (safety, legal, harassment, discrimination), include an escalation path.
  6. Suggest follow-up action if applicable.
  7. Deliver the tone-calibrated response ready for the user to post.

Prompt Templates

1. Response from Review (response_from_review)

Purpose: Generate a tone-calibrated reply from a single review. Input:

  • ${review_text} — The customer's original review
  • ${star_rating} — 1 through 5
  • ${platform} — Where the review was posted
  • ${verified_purchase} — (Optional) whether this is a verified buyer

Output: Complete reply text + tone classification + follow-up suggestion (if applicable).

2. Negative Review Diffuser (negative_review_diffuser)

Purpose: De-escalate an angry or very negative review with empathy and resolution. Input:

  • ${review_text} — The angry/negative review
  • ${specific_issues} — List the specific complaints raised
  • ${resolution_options} — What the business can actually do (refund, replacement, investigation)

Output: Empathetic de-escalation reply with: acknowledgment → specific action offered → private contact path → gratitude for feedback.

3. Positive Review Amplifier (positive_review_amplifier)

Purpose: Turn a 4–5★ review into a brand loyalty moment. Input:

  • ${review_text} — The positive review
  • ${what_they_loved} — Specific aspects they praised
  • ${upsell_opportunity} — (Optional) whether to include a gentle next-step invitation

Output: Warm thank-you reply that: references their specific praise → reinforces what makes your product great → optionally suggests related products or future engagement.

4. Response Policy Builder (response_policy_builder)

Purpose: Create an internal response policy document for a customer service team. Input:

  • ${brand_name} — Your brand
  • ${review_platforms} — Platforms where you receive reviews
  • ${response_sla} — Response time targets (e.g., "24 hours for negative, 48 for positive")
  • ${escalation_triggers} — What issues require manager attention

Output: Policy document with: response templates per star rating, tone guidelines, SLA by rating, escalation triggers, do-not-do list.

5. Multi-Language Response (multilanguage_response)

Purpose: Generate the same review response in multiple languages. Input:

  • ${review_text} — Original review (may be in any language)
  • ${response_text} — Your drafted response in your language
  • ${target_languages} — Languages needed (e.g., "English, Japanese, German")

Output: Multi-language response table: Language | Response Text | Cultural Note (if applicable).

Output Format

Every response includes:

TONE CLASSIFICATION: [Positive / Neutral / Negative / Angry]
STAR RATING: [★]
PLATFORM: [Platform]

RESPONSE:
[Complete reply text]

FOLLOW-UP: [Suggested next step, or "None required"]

Safety Rules

  • CRITICAL: NEVER generate fake positive reviews or impersonate a customer
  • CRITICAL: NEVER suggest review manipulation — removal requests, incentivized changes, or review gating
  • CRITICAL: Templates are for RESPONDING to existing reviews ONLY
  • NEVER write defensive, argumentative, or angry replies — maintain professionalism at all times
  • ALWAYS include an escalation path for complaints involving safety, legality, harassment, or discrimination
  • ALWAYS take conversations involving personal information or heated disputes to private channels (email, phone, DM)
  • NEVER share customer personal information in a public reply

Examples

Example 1: Negative Review Response

Input: Review="收到的衣服颜色和图片完全不一样,面料也很差,非常失望", Rating=2★, Platform="Taobao" Output: Classification=Negative. Response: "非常抱歉让您有这样的体验...我们已记录您反馈的颜色和面料问题...请您通过旺旺联系我们,为您办理退换货...感谢您的反馈,我们会优化产品图片的色差控制。"

Example 2: Positive Review Amplifier

Input: Review="第二次回购了,面霜保湿效果真的很好,冬天用刚刚好", Rating=5★, Platform="Amazon" Output: Classification=Positive. Response: "感谢您的持续支持!很高兴面霜在冬天帮到了您...我们最近也推出了同系列的精华,和面霜搭配效果更佳,欢迎了解...期待继续为您服务。"

Example 3: Safety Escalation

Input: Review="这个充电器用了一周就冒烟了,差点着火!", Rating=1★ Output: Immediately escalate: "您的安全是我们最关心的。请立即停止使用该产品,并通过[客服电话/邮箱]联系我们。我们将安排工程师检测并全额退款..."

Related Skills

Comments

Loading comments...