Magic Recap Template Library

Review the full ticket context (including title, conversation logs, internal notes, and resolution details). Generate a **CSAT survey message** with a **personalized opening note** that is **concise, clear, and human-sounding**. The note should briefly reference the customer's issue, what was accomplished, and kindly ask them to complete the survey.

Then, include the survey text in the exact **Markdown format** below, ensuring that the `{Ticket ID}` placeholder is correctly embedded in all URLs. **Do not deviate from this structure or formatting.**

**[Personalized Opening Note]**
> [A concise, human-sounding note tailored to the customer's ticket context that briefly mentions the issue and its resolution, and kindly asks them to complete the survey.]

Choose from the options below to rate your service experience:

[Service magic](https://feedback.smileback.io/{Ticket ID}/1/)
[Good service](https://feedback.smileback.io/{Ticket ID}/0/)
["I'm not mad; just disappointed"](https://feedback.smileback.io/{Ticket ID}/-1/)

Thank you!

Ensure the survey message follows this **Markdown format exactly** and that the opening note is **succinct, appropriately personalized, and includes a friendly request** to complete the survey.

Analyze the ticket details and generate an **Escalation Report** that provides a structured handover. Ensure that it is **concise, clear, and well-organized**.

Format:

**Issue Summary:** [Briefly describe the original problem]
**Troubleshooting Steps:** [Summarize actions already taken and results]
**Reason for Escalation:** [Clearly explain why escalation is required]
**Next Steps & Owner:** [Define what needs to happen next and who is responsible]

The report should be formatted using **bullet points for clarity**, avoid redundant information, and focus on **critical details for the escalation team**.

Review the full incident details, including **conversation logs, technician notes, timestamps, and resolution steps**. Generate an **Incident Report** with the following structured format:

**Incident Title:** [Summarize the issue in 5-10 words]
**Date & Time:** [Timestamp of incident detection]
**Affected Users/Systems:** [List impacted users, devices, or services]
**Description:** [Briefly describe what happened and the symptoms observed]
**Root Cause Analysis:** [Summarize identified causes, if known]
**Actions Taken:** [List steps taken to mitigate or resolve the issue]
**Final Resolution:** [Describe how the issue was resolved or next steps]

Ensure the **report is concise, uses bullet points where possible**, and is formatted clearly for review.

# Customer Experience (CX) Score Analysis Prompt

You are an AI assistant tasked with analyzing a customer service conversation and generating a Customer Experience (CX) Score. Your goal is to evaluate the overall customer satisfaction based on the entire conversation, considering both AI and human agent interactions.

## Scoring Framework

Generate a score from 1-5 where:
- **1-3**: Negative experience
- **4-5**: Positive experience

The final CX Score percentage = (Number of positive ratings [4s and 5s] / Total ratings) × 100

## Key Evaluation Criteria

Analyze the conversation across three primary dimensions:

### 1. Customer Sentiment
- **Tone Analysis**: Evaluate the customer's language and emotional tone throughout the conversation
- **Sentiment Progression**: Track how the customer's mood evolves from start to finish
- **Frustration Indicators**: Look for signs of irritation, repeated questions, or escalating language
- **Satisfaction Signals**: Identify expressions of gratitude, understanding, or positive acknowledgment

### 2. Resolution Status
- **Issue Identification**: List all problems or questions raised by the customer
- **Resolution Assessment**: Determine which issues were fully resolved, partially addressed, or left unresolved
- **Clarity of Solutions**: Evaluate whether solutions were clearly communicated and understood
- **Follow-up Needs**: Assess if the customer needs additional assistance

### 3. Service Quality
- **Response Timeliness**: Consider speed of responses relative to complexity
- **Knowledge Demonstration**: Evaluate accuracy and depth of information provided
- **Communication Clarity**: Assess how well agents explained solutions
- **Professionalism**: Rate courtesy, empathy, and helpfulness
- **Proactive Support**: Note any anticipation of customer needs

## Scoring Guidelines

### Score 5 - Exceptional Experience
- All issues completely resolved
- Customer expresses clear satisfaction
- Service exceeded expectations
- Proactive and empathetic support

### Score 4 - Positive Experience
- Main issues resolved satisfactorily
- Generally positive customer tone
- Good service quality with minor room for improvement
- Customer needs were met

### Score 3 - Neutral/Mixed Experience
- Some issues resolved, others pending
- Mixed customer sentiment
- Adequate but not outstanding service
- Basic needs met with some friction

### Score 2 - Negative Experience
- Key issues unresolved or poorly handled
- Customer frustration evident
- Service quality below expectations
- Multiple pain points in the interaction

### Score 1 - Very Poor Experience
- Critical issues unresolved
- High customer frustration/anger
- Poor service quality
- Customer explicitly dissatisfied

## Output Format

Provide your analysis in the following structure:
CX SCORE: [1-5]
CUSTOMER SENTIMENT ANALYSIS:

Initial sentiment: [Positive/Neutral/Negative]
Final sentiment: [Positive/Neutral/Negative]
Key emotional indicators: [List specific phrases or behaviors]

RESOLUTION STATUS:

Issues identified: [List each issue]
Resolution rate: [X out of Y issues resolved]
Unresolved items: [List if any]

SERVICE QUALITY ASSESSMENT:

Response time: [Excellent/Good/Average/Poor]
Knowledge level: [Excellent/Good/Average/Poor]
Communication: [Excellent/Good/Average/Poor]
Overall helpfulness: [Excellent/Good/Average/Poor]

SCORE JUSTIFICATION:
[2-3 sentences explaining why this specific score was assigned, highlighting the most impactful factors]
KEY MOMENTS:

Positive: [List standout positive moments]
Negative: [List friction points or failures]

IMPROVEMENT OPPORTUNITIES:
[1-2 specific suggestions for better handling similar conversations]

## Special Considerations

1. **Multi-participant conversations**: When both AI and human agents are involved, evaluate the overall experience rather than individual performances
2. **Context sensitivity**: Consider the complexity and urgency of the customer's issues
3. **Cultural/language factors**: Be aware of communication style differences
4. **Technical limitations**: Recognize when issues are beyond agent control (system outages, policy restrictions)

## Exclusions

Do not score if:
- The conversation appears to be spam
- There are fewer than 2 customer messages
- There are fewer than 2 agent responses
- The conversation lacks sufficient information for assessment

**Remember**: Focus on the complete customer journey and overall satisfaction, not just the final outcome. A well-handled difficult situation can still result in a positive score if the service quality was exceptional.

Analyze the conversation logs and **identify customer sentiment** based on tone, keywords, and phrasing. Provide a **sentiment summary** that includes:

- **Overall Sentiment:** [Positive / Neutral / Negative]
- **Key Indicators:** [Mention key phrases or words that influenced this sentiment]
- **Suggested Follow-Up Action:** [If sentiment is negative, recommend a follow-up approach]

Ensure the summary is **concise, to the point, and highlights actionable insights**.

Review the change request and generate a **structured Change Management Log**:

**Change Request Title:** [Summarize the change in a few words]
**Requested Date & By Whom:** [List request date and requester]
**Description of Change:** [Summarize what change was requested and why]
**Approval Status:** [Approved / Pending / Denied – Include details if needed]
**Actions Taken:** [List steps taken, including timestamps and responsible parties]
**Outcome & Notes:** [Summarize the result and any key observations]

Ensure the log is **structured for easy reference and compliance tracking**.

Analyze the ticket and generate **Compliance & Audit Notes** that capture key information for compliance tracking.

Format:

**Compliance Area:** [Security / Data Handling / Access Control / etc.]
**Description:** [Summarize the compliance-related issue or event]
**Actions Taken:** [List compliance steps taken, including any approvals or documentation updates]
**Notes & Next Steps:** [Any additional actions required for compliance tracking]

Ensure the summary is **clear, factual, and structured for easy reference**.

Review the ticket details and generate a **Knowledge Base Draft** in a structured format.

# **[Issue Title]**
### **Issue Summary**
[A concise summary of the problem users face]

### **Symptoms**
- [List key symptoms users may experience]

### **Resolution Steps**
1. **[Step 1]** – [Detailed explanation]
2. **[Step 2]** – [Detailed explanation]
3. **[Step 3]** – [Detailed explanation]

### **Additional Notes**
- [Any additional context, workarounds, or related articles]

---

Ensure the **draft is structured clearly** and uses **consistent formatting for easy readability**.