| name | support-deflection-miner |
| description | Analyze support tickets, helpdesk exports, issue subjects, customer questions, or support thread summaries to find repeated themes, duplicate-looking requests, documentation gaps, automation candidates, and support-deflection opportunities. Use when Codex needs to convert support noise into prioritized KB, product, or workflow improvements without inventing customer facts. |
Support Deflection Miner
Workflow
- Identify source type, date range, product area, customer segment, and support goal.
- Group repeated questions, symptoms, ticket subjects, and workaround requests.
- Separate documentation gaps, product friction, policy confusion, and automation opportunities.
- Prioritize by repetition, customer impact, confidence, and actionability.
- Produce proposed KB articles, product feedback, macros, automation candidates, and caveats.
Script-Assisted Workflow
When given ticket subjects in CSV or text form, run scripts/mine_support_themes.py --input <path> first. Use --json when structured theme counts are needed. The helper surfaces repeated text patterns; Codex must still judge whether a theme is a KB gap, product issue, or workflow candidate.
Output Standard
Use this structure by default:
- Deflection Summary: source scope, top repeated issues, and confidence.
- Top Themes: theme, evidence, count, likely category, and caveats.
- Duplicate-Looking Requests: repeated subjects or near-repeated issue language.
- KB Opportunities: proposed article titles, audience, source evidence, and missing facts.
- Product / Process Signals: friction points requiring product, policy, or workflow review.
- Automation Candidates: macro, bot, form, routing, or script opportunities.
- Questions To Resolve: missing context needed before publishing or automating.
Rules
- Do not invent customer names, ticket counts beyond the provided input, root causes, SLA impact, or product commitments.
- Mark keyword clusters as heuristic, not definitive taxonomy.
- Keep customer-sensitive data out of examples unless already sanitized.
- Prefer small, reviewable deflection actions over broad automation claims.
References
Read references/deflection-rubric.md when deciding whether a repeated support theme should become a KB article, product bug, macro, or automation candidate.
