Passed 19 of 22 test cases with the skill, and 17 of 22 without it.
Without skill77%
With skill86%
Checks it passed that the model alone missed
The output includes a generalized, reusable template version extracted from the drafted response.
The package specifies an optimal response timing window based on platform algorithms and urgency.
The response prompts the user to set up a brand first or offers to proceed using standard defaults.
3 fixed16 passed either way2 still failing1 broke
Pass-rate lift
+9 pp
Tokens
+64%
Agent turns
0%
Response time
+2%
Model gemini-3.6-flashMeasured 8/17/2026v119 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 3
Prompt
A 1-star urgent review on Google claims an overcharge on a rental car deposit. Generate the complete response package.
What was checked, without and with the skill
Checked
Without skill
With skill
The package specifies an optimal response timing window based on platform algorithms and urgency.
Missed
Passed
case-18 · gemini-3.6-flash · 8/17/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md54 linessingle file
/digital-marketing-pro:review-response
Purpose
Generate professional, brand-aligned review responses for positive, neutral, and negative reviews across any platform. Ensures every response maintains brand voice, addresses the reviewer's specific points, and follows best practices for reputation management and customer recovery.
Input Required
The user must provide (or will be prompted for):
Review text: The full text of the review to respond to
Rating: Star rating (1-5 stars)
Platform: Where the review was posted (Google, Yelp, G2, Capterra, Trustpilot, Amazon, TripAdvisor, App Store, etc.)
Reviewer name: Display name of the reviewer (optional — for personalization)
Specific issue mentioned: Key complaint, praise, or topic raised in the review (optional — for targeted response)
Business context: Any internal context about the situation — was the issue resolved, is there a known product bug, was there a service failure (optional — helps craft an accurate response)
Batch mode: If responding to multiple reviews, provide them as a set for consistent tone and varied language
Response speed requirement: Whether the review needs an urgent response (crisis situation) or standard turnaround
Internal resolution status: Whether the issue has been fixed, is in progress, or is unresolved (for negative reviews — helps determine what to promise)
Process
Load brand context: Read ~/.claude-marketing/brands/_active-brand.json for the active slug, then load ~/.claude-marketing/brands/{slug}/profile.json. Apply brand voice, compliance rules for target markets (skills/context-engine/compliance-rules.md), and industry context. Also check for guidelines at ~/.claude-marketing/brands/{slug}/guidelines/_manifest.json — if present, load restrictions and relevant category files. Check for custom templates at ~/.claude-marketing/brands/{slug}/templates/. Check for agency SOPs at ~/.claude-marketing/sops/. If no brand exists, ask: "Set up a brand first (/digital-marketing-pro:brand-setup)?" — or proceed with defaults.
Apply brand voice settings: Load voice-and-tone guidelines and any channel-specific style rules for the review platform — review responses often require a warmer, more personal tone than other brand communications
Classify review sentiment and severity: Categorize as positive (4-5 stars), neutral (3 stars), or negative (1-2 stars) — further classify negative reviews by severity level: minor complaint, service failure, product defect, or safety/legal issue
For negative reviews: Acknowledge the specific concern by name, express genuine empathy without generic platitudes, take responsibility where appropriate, offer a concrete resolution path with specifics, and move the conversation offline with a direct contact method (email or phone)
For positive reviews: Express sincere gratitude, reinforce the specific aspect the reviewer praised, add a personal or humanizing touch, and encourage continued engagement — mention related products, services, or referral programs where natural
For neutral reviews: Acknowledge the balanced feedback, address any specific concerns raised with actionable detail, highlight relevant brand strengths without being defensive or dismissive, and invite further dialogue to improve their experience
Check brand guidelines for approved response language: Verify the response against any restricted terms, required disclosures, legal disclaimers, or mandated response elements in the brand guidelines
Apply platform conventions: Adjust response length, formatting, and tone for platform norms — Google (concise), Yelp (conversational), G2 (professional), TripAdvisor (hospitality-focused), etc.
Score response for brand voice alignment: Evaluate the drafted response against brand voice parameters — tone, formality, warmth, and personality — and adjust until the response sounds authentically on-brand
Check for common pitfalls: Ensure the response avoids defensiveness, blame-shifting, over-promising, disclosing private information, or using repetitive language across multiple review responses
Optimize for SEO where applicable: On platforms where responses are indexed (Google, Yelp), naturally incorporate relevant keywords and business name without sounding forced
Generate batch variations: If responding to multiple similar reviews, vary the language, structure, and opening to avoid templated-sounding responses that damage authenticity
Output
A structured review response package containing:
Ready-to-post review response: Primary response tailored to the platform's character limits, conventions, and audience expectations