# Support Escalator

> Reads an Intercom conversation, reproduces the bug, and files a linked Jira issue with repro steps.

- Skill: `fadil369/support-escalator` (Agent Skill)
- Install (CLI): `npx skillmds@latest add fadil369/support-escalator`
- Raw SKILL.md: https://api.skillmd.com/api/skills/fadil369/support-escalator/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: fadil369 (https://skillmd.com/u/fadil369)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/fadil369/support-escalator

---


# Support-to-Eng Escalator

Bridge support and engineering. Given an Intercom conversation, pull context, attempt reproduction, file a Jira issue, and notify the team — end to end.

## How to invoke

Provide:
- **Intercom conversation ID** (required)
- **Jira project key** for the engineering project (e.g. `ENG`, `BACKEND`)
- **Slack channel** for support notifications (e.g. `#support-escalations`)

## Process

### 1. Pull the conversation from Intercom
Extract:
- Customer name, plan tier, and account ID
- Environment details (browser, OS, app version, region)
- Any attached logs, error messages, or screenshots
- Support rep's notes and tags
- Timeline of the issue

### 2. Attempt reproduction in the session environment
- Follow the steps described by the customer exactly
- Capture the minimal triggering command, request payload, or UI flow
- If repro succeeds: record the exact reproducing case
- If repro fails: document every variation attempted and why each failed

**Never file a vague "cannot reproduce" issue.** If you can't repro, list what you tried.

### 3. Create the Jira issue
Fields to set:
- **Summary**: concise, specific (e.g. "Checkout fails with 500 when cart has >10 items on Pro plan")
- **Description**: customer context, minimal repro steps, environment, suspected component (from code search)
- **Links**: Intercom conversation URL, any relevant GitHub commits or PRs
- **Labels**: `customer-reported`, plan tier, component
- **Priority**: based on plan tier + user impact (P1 for Enterprise/data-loss, P2 for paying users, P3 for free)

### 4. Notify the support Slack channel
Post a single message:
- Conversation escalated ✅
- Jira issue link
- One-sentence severity assessment
- Tag the on-call engineer if P1

### 5. Update the Intercom conversation
- Add an internal note with the Jira link
- Mark the conversation as escalated
- Set the next follow-up reminder

## Output summary

After completing all steps, reply with:
```
Escalation complete:
- Jira: [PROJ-XXX](link) — [Summary]
- Slack: notified #channel
- Intercom: internal note added, marked escalated
- Repro: [success / failed — tried: X, Y, Z]
```

## Requirements

- **Intercom MCP** required. Configure at `https://mcp.intercom.com/mcp`.
- **Atlassian MCP** required for Jira. Configure at `https://mcp.atlassian.com/v1/mcp`.
- **Slack MCP** required for notifications. Configure at `https://mcp.slack.com/mcp`.