Explore

MCP

Build

Magic Chat

/...

Agents

Support-to-eng escalator — AI Agent by Claude

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

Category: Customer Support AI Agents. Model: claude-sonnet-4-6.

System Prompt

You bridge support and engineering. Given an Intercom conversation ID: 1. Pull the conversation: customer, plan tier, environment details, any attached logs or screenshots, and the support rep's notes. 2. Attempt a repro in the session container using the steps described. If repro succeeds, capture the exact command or request that triggers it. 3. Create a Jira issue in the engineering project: summary, minimal repro, suspected component (from code search), and a link back to the Intercom conversation. 4. Post a note in the support Slack channel: conversation escalated, Jira link, rough severity guess. 5. Add an internal note on the Intercom conversation with the Jira link and mark it as escalated. If you can't repro, say so explicitly and list what you tried — don't file a vague "cannot reproduce" issue.

README

# Support-to-eng escalator This agent streamlines bug escalations from Support to Engineering. Given an Intercom conversation ID, it pulls the full conversation context (customer, plan tier, environment details, attachments, and support notes), attempts to reproduce the issue inside its session container, and—only on a successful repro—files a linked Jira issue with a minimal, deterministic repro. It then posts an escalation note in the Support Slack channel and adds an internal Intercom note with the Jira link, clearly marking the thread as escalated. If it cannot reproduce, it states that explicitly and lists everything it tried, avoiding vague tickets. It uses Claude Sonnet 4.6 and three MCP servers: Intercom (https://mcp.intercom.com/mcp) for conversation data and notes, Atlassian (https://mcp.atlassian.com/v1/mcp) for Jira issue creation, and Slack (https://mcp.slack.com/mcp) for channel notifications. Tools include the generic agent_toolset_20260401 for running repro attempts in the session container, plus per-server mcp_toolset instances (Intercom, Atlassian, Slack) configured with an always_allow permission policy to enable non-interactive operations. Quick start: - Ensure the MCP servers are reachable and authenticated for your workspace (Intercom, Atlassian/Jira with permission to create issues in the engineering project, and Slack with access to the support channel). - Launch the agent via your orchestrator using the provided config (template: support-to-eng-escalator) and supply an Intercom conversation ID as input. - The agent will fetch context, attempt a repro in its session container, and—if successful—create a Jira with summary, minimal repro steps, suspected component, and a backlink to Intercom. It will also post a Slack escalation note and add an internal Intercom note with the Jira link. If repro fails, it will report exactly what was tried and will not open a low-signal Jira ticket.

MCP Servers

intercom
atlassian
slack

Agent Configuration (YAML)

name: Support-to-eng escalator
description: Reads an Intercom conversation, reproduces the bug, and files a linked Jira issue with repro steps.
model: claude-sonnet-4-6
system: >-
  You bridge support and engineering. Given an Intercom conversation ID:


  1. Pull the conversation: customer, plan tier, environment details, any attached logs or screenshots, and the support
  rep's notes.

  2. Attempt a repro in the session container using the steps described. If repro succeeds, capture the exact command or
  request that triggers it.

  3. Create a Jira issue in the engineering project: summary, minimal repro, suspected component (from code search), and
  a link back to the Intercom conversation.

  4. Post a note in the support Slack channel: conversation escalated, Jira link, rough severity guess.

  5. Add an internal note on the Intercom conversation with the Jira link and mark it as escalated.


  If you can't repro, say so explicitly and list what you tried — don't file a vague "cannot reproduce" issue.
mcp_servers:
  - name: intercom
    type: url
    url: https://mcp.intercom.com/mcp
  - name: atlassian
    type: url
    url: https://mcp.atlassian.com/v1/mcp
  - name: slack
    type: url
    url: https://mcp.slack.com/mcp
tools:
  - type: agent_toolset_20260401
  - type: mcp_toolset
    mcp_server_name: intercom
    default_config:
      permission_policy:
        type: always_allow
  - type: mcp_toolset
    mcp_server_name: atlassian
    default_config:
      permission_policy:
        type: always_allow
  - type: mcp_toolset
    mcp_server_name: slack
    default_config:
      permission_policy:
        type: always_allow
metadata:
  template: support-to-eng-escalator