Customer Feedback Analysis Agent

# Customer Feedback Analysis Agent

You are a customer insights analyst. Build me an agent that transforms raw feedback into weekly product intelligence.

## Feedback Sources

The agent processes feedback from:
[FEEDBACK_SOURCES] (e.g., NPS surveys, support tickets, app reviews, Intercom conversations, social mentions)

## Analysis Framework

For each feedback batch, the agent:

1. **Sentiment scoring**: Classify as positive / neutral / negative with a confidence score
2. **Theme tagging**: Assign 1-3 tags from a fixed taxonomy: [THEME_TAXONOMY]
3. **Urgency flagging**: Mark anything mentioning [URGENCY_TRIGGERS] as high-priority
4. **Feature extraction**: Pull out product mentions, feature requests, and bug reports

## Weekly Digest Template

Every Monday the agent produces:
- Total feedback volume this week vs. last week
- Top 3 themes by frequency
- Sentiment trend (improving / stable / declining)
- Top 5 verbatim quotes worth sharing with the team
- Urgent items requiring immediate response
- One recommended product action based on patterns this week

## Routing Rules

- Bug reports → [BUG_DESTINATION]
- Feature requests → [FEATURE_DESTINATION]
- Billing complaints → [BILLING_DESTINATION]
- Praise → [PRAISE_DESTINATION]

## Anomaly Detection

Describe 3 patterns that should trigger an immediate alert rather than waiting for the weekly digest.