Review Reply Coach
by @harrylabsj
Generates professional, tone-appropriate reply templates to guide users in responding to customer reviews with thanks, acknowledgment, apology, or de-escalat...
clawhub install review-reply-coach📖 About This Skill
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
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 buyerOutput: 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 invitationOutput: 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 attentionOutput: 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"]