order-troubleshooting

$npx mdskill add langchain-ai/langsmith-agent-lifecycle-workshop/order-troubleshooting

1. Confirm the customer is verified. If not, ask them to verify before you look anything up. 2. Ask the `order-specialist` for the order's status, order date, ship date, and items. 3. Match the situation: - `Processing` for more than 3 days: apologize, say it is being prioritized, and offer cancellation (per the Escalation Policy). - `Shipped`: give the ship date and the expected delivery window from the Shipping Guide (ask the `product-policy-specialist`). - `Delivered` but the customer says it didn't arrive: escalate to a human (possible lost package). - Wrong or missing item: compare the order items with what the customer describes, then offer a return or replacement. 4. Never guess a tracking status the tools didn't return.

SKILL.md

.github/skills/order-troubleshootingView on GitHub ↗
---
name: order-troubleshooting
description: Diagnose late, missing, or wrong TechHub orders using order status, ship dates, and items. Use when a customer asks where an order is or reports a problem with what arrived.
---

# Order troubleshooting

## Steps

1. Confirm the customer is verified. If not, ask them to verify before you look anything up.
2. Ask the `order-specialist` for the order's status, order date, ship date, and items.
3. Match the situation:
   - `Processing` for more than 3 days: apologize, say it is being prioritized, and offer cancellation (per the Escalation Policy).
   - `Shipped`: give the ship date and the expected delivery window from the Shipping Guide (ask the `product-policy-specialist`).
   - `Delivered` but the customer says it didn't arrive: escalate to a human (possible lost package).
   - Wrong or missing item: compare the order items with what the customer describes, then offer a return or replacement.
4. Never guess a tracking status the tools didn't return.

## Output

- Current status in one line.
- What happens next, and when.
- Any action you took or are escalating.

More from langchain-ai/langsmith-agent-lifecycle-workshop