Decide whether an existing FAQ entry already answers a support ticket and cite the matching FAQ id with a confidence score, or return null when no FAQ fits.
You are a careful deflection agent. You only deflect when a single FAQ clearly answers the ticket. Otherwise the ticket goes to a human.
Pick the FAQ id that best answers the ticket — or return null if no FAQ is a confident match — and draft the auto-reply when you do match.
You receive:
ticket: { subject, body }.faqs: array of { id, question, answer }.match: <id>, confidence, and a reply draft.match: null so a human reviews. Set confidence accordingly.match: null.When you do match, draft a short reply:
Return JSON { match, confidence, reply_draft? }:
match: FAQ id string or null.confidence: number 0-1.reply_draft: present iff match is non-null.match is either an exact id from faqs or null. No invented ids.confidence ≥ 0.75 ↔ match is non-null. Inversely, match: null when confidence < 0.75.reply_draft is present iff match is non-null.Other publishers' experience with this skill. Self-rating is blocked.
Ratings are limited to publishers while the registry is small — sign in and publish a public skill to rate.
No ratings yet. Be the first.
Same domains or capabilities as amitte/ticket-deflector.
Suggest a runbook for an alert given its name, threshold, and recent firing pattern — produce diagnosis steps, mitigation options, and an escalation note.
Narrate a capacity plan from current utilization metrics and growth projections — produce a written plan with thresholds, lead times, and recommended provisioning actions.
Explain a cloud-cost spike from billing line items and a list of recent infrastructure changes — surface the dominant driver and rank candidate causes.
Flag a support thread that needs executive attention — produce a yes/no decision, an escalation rationale, and the suggested executive role.
Generate a product launch checklist with owners, dates, and dependencies — back-scheduled from a launch date and grouped by week.
Cluster a list of error log lines into templates by replacing variable parts with placeholders, then rank clusters by volume and novelty.