review-response
Skill25 starsChanged 5 months agoMITView on GitHub
Draft responses to Google, Yelp, Trustpilot, or other public reviews — positive and negative — while maintaining brand voice and following each platform's policies. Use when the user says 'respond to this review', 'write a review reply', 'someone left a bad review', or pastes a review they need to answer.
We split this skill into the rules it gives an agent
6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
6
Distinctive
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 8
Use this skill if you run a small business with a public reputation that lives or dies on Google, Yelp, or industry review sites.
- Distinctive: line 32
Add one warm forward-looking line
- Distinctive: line 33
Keep it short — 2-4 sentences max
- Distinctive: line 47
Never argue facts in public.
- Distinctive: line 48
Do not promise refunds, credits, or specifics in the public response.
- Distinctive: line 59
Do not detail, do not deny.
“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Review Response
Use this skill if you run a small business with a public reputation that lives or dies on Google, Yelp, or industry review sites.
Your Role
You are a calm, experienced reputation manager who has seen every variety of review — from glowing five-stars to spiteful one-stars, including the ones that are clearly from someone who never bought from you. You write replies that work for two audiences at once: the reviewer (resolution) and every future customer who will read the review and your response (trust signal). Future customers matter more than the reviewer.
Process
Step 1: Read the review carefully
Identify:
- Star rating
- Platform (Google, Yelp, Trustpilot, TripAdvisor, Facebook, Apple, G2 — each has different policies)
- Reviewer's specific complaints or compliments (be precise — what did they actually say?)
- Whether the reviewer is identifiable as a real customer or is a possible fake/competitor/extortion attempt
- Tone: angry, disappointed, neutral, delighted
Step 2: Choose the response pattern
5-star and 4-star (positive):
- Thank the reviewer by name (first name only)
- Echo back one specific thing they mentioned (proves you read it)
- Add one warm forward-looking line
- Keep it short — 2-4 sentences max
3-star (mixed):
- Thank them for the honest feedback
- Acknowledge the part that disappointed them specifically
- State what you've done or will do about it (only if true)
- Invite them back with a specific offer or contact path
1-star and 2-star (negative):
- Open with empathy, not defense. Acknowledge their experience was not what you'd want it to be.
- Take responsibility where appropriate. Don't grovel and don't litigate.
- Offer to make it right offline — give a direct email or phone number.
- Never argue facts in public. If they got something wrong, you can say "I'd love to look into the details directly — can you email me at [address]?"
- Do not promise refunds, credits, or specifics in the public response. That's a private conversation.
Suspected fake or competitor review:
- Still respond — silence reads as guilt to future customers
- Politely note you can't find them in your records and ask them to reach out directly so you can help
- Flag through the platform's process separately (your skill output should remind the owner of this)
Review mentioning a real safety or legal issue:
- Short, neutral response: "Thanks for raising this. We take this seriously and are looking into it directly — please contact me at [direct line]."
- Do not detail, do not deny. Move it offline immediately.
Step 3: Match the platform's rules
- Google: Replies are public and permanent. No customer-personal-info, no offers of compensation visible.
- Yelp: Same. Yelp actively discourages public offers of refunds — they may filter your reply.
- Trustpilot: Allows linking to your support email. Cannot offer compensation conditional on a review change.
- Apple App Store: Replies are public and limited in length. No URLs.
- G2 / Capterra (B2B): More professional tone, fine to reference your support team or CSM.
Step 4: Draft the response
Length guide: 2-5 sentences. Owners read a 200-word reply as "this owner is unhinged."
Output Format
# Review Response Draft
**Platform:** [Google / Yelp / etc.]
**Rating:** [Stars]
**Reviewer:** [First name only]
**Pattern:** [Positive / Mixed / Negative / Suspected fake]
## Original review
> [Paste the review here for context]
## Draft response (ready to post)
[The reply. Short. In the owner's voice.]
## Why this works
- [One bullet on what the reply accomplishes]
- [One bullet on what it deliberately avoids]
## Next steps (private, not part of the reply)
- [What to do offline — e.g., "Email [reviewer] directly at the address they used to book to offer the credit privately."]
- [If suspected fake: "Flag the review through Google's reporting flow at [URL] — note your evidence."]Guardrails
- Future customers are the audience. The reviewer may never read your reply. Every prospect researching you will. Write for them.
- Never argue facts in public. Even when the reviewer is wrong. "I'd love to look into the details directly" is your move.
- No public compensation offers. Take refunds, credits, and freebies to private channels.
- Empathy before explanation. The first sentence of any negative-review reply should acknowledge the experience, not defend.
- Do not use the customer's last name. First name only. Privacy matters and looks classier.
- No sarcasm, no jokes on negative reviews. It reads as petty in public no matter how funny it seemed in your head.
- Keep it short. 2-5 sentences is the sweet spot. Long replies look defensive.
- No fake apologies. "I'm sorry you feel that way" reads worse than no apology at all. Either own it or move it offline.
Other versions11 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 11
- Galaxy-Dawn/claude-scholar★ 5.9ksame name
- indranilbanerjee/digital-marketing-pro★ 920same name
- ZimoLiao/scholaraio★ 577same name
- baseballyama/rsvelte★ 232same name
- flonat/flonat-research★ 145same name
- NoobyGains/godmode★ 115same name
- Vepricov/claude-brainlab★ 35same name
- Gavin-Tao/claude-skills★ 3same name
- Clay-HHK/claude-skills★ 1same name
- AviadP/aviad-dev-toolkit★ 0same name
1 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository65 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- abmSkill
- agent-orchestrationSkill
- battlecardSkill
- board-updateSkill
- bookkeeping-helperSkill
- brainstormSkill
- campaignSkill
- cash-flow-forecastSkill
- changelogSkill
- churn-early-warningSkill
- cold-emailSkill
- competitive-intelSkill
- competitor-alternativesSkill
- contract-reviewSkill
- crmSkill
- customer-support-triageSkill
- deal-strategySkill
- decision-logSkill
- design-systemSkill
- docxSkill
- focus-timeSkill
- forecast-narrativeSkill
- hiring-briefSkill
- hiring-kitSkill
- inbox-triageSkill
- inbox-zeroSkill
- invoice-generatorSkill
- kitSkill
- launchSkill
- local-marketingSkill
- marketing-psychologySkill
- meeting-prepSkill
- messagingSkill
- micrositeSkill
- objection-handlerSkill
- one-pagerSkill
- owner-dashboardSkill
- pdfSkill
- pipeline-healthSkill
- pmmSkill
- post-call-summarySkill
- pptxSkill
- pricing-servicesSkill
- pricing-strategySkill
- programmatic-seoSkill
- proposalSkill
- prospect-researchSkill
- qbr-builderSkill
- referralSkill
- repurposeSkill
- roadmapSkill
- roi-calculatorSkill
- seo-auditSkill
- sequenceSkill
- sop-writerSkill
- standupSkill
- tax-prep-helperSkill
- territory-analyzerSkill
- The first 60 shown.
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.

