Alert your support lead in Slack when Gorgias tickets approach SLA deadline using a Claude Code skill

mkdir -p .claude/skills/sla-alert/{templates,references}

.claude/skills/sla-alert/
├── SKILL.md                          # workflow guidelines + config
├── templates/
│   └── slack-alert.md                # Block Kit template for SLA warnings
└── references/
    └── gorgias-tickets-api.md        # Gorgias API + SLA calculation

---
name: sla-alert
description: Check open Gorgias tickets for approaching SLA deadlines and post warnings to Slack so the team lead can reassign or escalate before a breach occurs.
disable-model-invocation: true
allowed-tools: Bash, Read
---
 
## Goal
 
Poll open Gorgias tickets that haven't received a first agent response, calculate how much SLA time remains for each, and post a Slack alert for any ticket approaching its deadline (default: 75% of SLA window elapsed).
 
## Configuration
 
Read these environment variables:
 
- `GORGIAS_DOMAIN` — your Gorgias subdomain (required)
- `GORGIAS_EMAIL` — Gorgias account email for API auth (required)
- `GORGIAS_API_KEY` — Gorgias API key (required)
- `SLACK_BOT_TOKEN` — Slack bot token with chat:write (required)
- `SLACK_CHANNEL_ID` — Slack channel ID starting with C (required)
 
Default SLA targets (in minutes): email=240, chat=15, contact-form=240. Default warn threshold: 75% elapsed. The user may request different values.
 
## Workflow
 
1. Validate that all required env vars are set. If any are missing, print which ones and exit.
2. Fetch open tickets from Gorgias. See `references/gorgias-tickets-api.md` for the endpoint and SLA calculation logic.
3. Filter to tickets with no agent reply (no message with `source.type == "helpdesk"` or `sender.type == "agent"`).
4. For each unanswered ticket, calculate elapsed time since creation, look up the SLA target for the ticket's channel, and determine if the warn threshold has been crossed.
5. Track already-alerted ticket IDs in a local JSON file to avoid duplicate warnings.
6. Post a Slack alert for at-risk tickets using the template in `templates/slack-alert.md`.
7. Print a summary of alerts sent.
 
## Important notes
 
- This calculates SLA using wall-clock time, not business hours. If your SLA counts only business hours, mention this to the user and suggest they define business hours in the SLA targets.
- `SLACK_CHANNEL_ID` must be the channel ID (starts with `C`), not the channel name. The `chat.postMessage` API requires the ID.
- The Slack bot must be invited to the target channel or posting will fail with `not_in_channel`.
- Clean up the alerted tickets file periodically (remove entries older than 24 hours) to prevent unbounded growth.
- Use the `requests` library for Gorgias and `slack_sdk` for Slack. Install with pip if needed.

# SLA Warning Slack Template
 
Use this Block Kit structure for each SLA warning alert.
 
## Block Kit JSON
 
```json
{
  "channel": "<SLACK_CHANNEL_ID>",
  "text": "SLA warning: <ticket_subject> — <status>",
  "blocks": [
    {
      "type": "header",
      "text": {
        "type": "plain_text",
        "text": "SLA Warning — Ticket Approaching Deadline"
      }
    },
    {
      "type": "section",
      "fields": [
        { "type": "mrkdwn", "text": "*Customer*\n<customer_name>" },
        { "type": "mrkdwn", "text": "*Channel*\n<channel>" },
        { "type": "mrkdwn", "text": "*Time Remaining*\n<remaining_or_breached>" },
        { "type": "mrkdwn", "text": "*SLA Target*\n<sla_minutes> min" },
        { "type": "mrkdwn", "text": "*Subject*\n<subject>" }
      ]
    },
    {
      "type": "actions",
      "elements": [
        {
          "type": "button",
          "text": { "type": "plain_text", "text": "Open in Gorgias" },
          "url": "https://<GORGIAS_DOMAIN>.gorgias.com/app/ticket/<ticket_id>",
          "style": "danger"
        }
      ]
    }
  ]
}
```
 
## Notes
 
- Use `danger` style on the button for visual urgency.
- For `<remaining_or_breached>`: show "BREACHED" if time remaining <= 0, otherwise show "X min remaining".
- Send via `slack_sdk` `chat.postMessage` with the `SLACK_BOT_TOKEN`.

# Gorgias Tickets API + SLA Calculation
 
## Authentication
 
All requests use HTTP Basic Auth:
- Username: `GORGIAS_EMAIL`
- Password: `GORGIAS_API_KEY`
- Base URL: `https://{GORGIAS_DOMAIN}.gorgias.com/api`
 
## List open tickets
 
**Request:**
 
```
GET /api/tickets?status=open&limit=100
```
 
**Response shape:**
 
```json
{
  "data": [
    {
      "id": 18201,
      "subject": "Shipping delay on order #4892",
      "status": "open",
      "channel": "email",
      "created_datetime": "2026-03-05T10:00:00+00:00",
      "requester": {
        "id": 678,
        "name": "Emily Tran",
        "email": "emily@example.com"
      },
      "messages": [
        {
          "id": 99001,
          "body_text": "Customer message here...",
          "source": { "type": "email" },
          "sender": { "type": "customer" }
        }
      ]
    }
  ]
}
```
 
## Determining if a ticket has an agent reply
 
Check each message in the `messages` array:
- `source.type == "helpdesk"` indicates an agent or system message
- `sender.type == "agent"` also indicates an agent message
- If any message matches either condition, the ticket has been replied to
 
## SLA Calculation Logic
 
Default SLA targets (in minutes):
 
| Channel | SLA (minutes) |
|---------|--------------|
| email | 240 (4 hours) |
| chat | 15 |
| contact-form | 240 |
| default | 240 |
 
**Steps:**
1. Parse `created_datetime` as UTC
2. Calculate elapsed minutes: `(now_utc - created_datetime) / 60`
3. Look up SLA target for the ticket's `channel`
4. Calculate remaining: `sla_target - elapsed`
5. Calculate percent elapsed: `elapsed / sla_target`
6. If percent elapsed >= 0.75 (warn threshold), the ticket is at risk
7. If remaining <= 0, the ticket has breached
 
**Note:** This uses wall-clock time. If your SLA only counts business hours, you need additional logic to subtract non-business hours from elapsed time.

/sla-alert

Checking open tickets for SLA risk...
Found 34 open ticket(s)
 
  #18201  "Shipping delay on order #4892"    42m left
    -> Slack alert sent
  #18195  "Wrong item received"              BREACHED
    -> Slack alert sent
  #18210  "Return label request"             15m left
    -> Already alerted (skipping)
 
Done. 2 alert(s) sent.

# Run every 15 minutes during business hours
*/15 8-18 * * 1-5 cd /path/to/your/project && claude -p "Run /sla-alert" --allowedTools 'Bash(*)' 'Read(*)'

name: SLA Breach Monitor
on:
  schedule:
    - cron: '*/15 * * * *'  # Every 15 minutes
  workflow_dispatch: {}
jobs:
  alert:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: anthropics/claude-code-action@v1
        with:
          prompt: "Run /sla-alert"
          allowed_tools: "Bash(*),Read(*)"
        env:
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
          GORGIAS_DOMAIN: ${{ secrets.GORGIAS_DOMAIN }}
          GORGIAS_EMAIL: ${{ secrets.GORGIAS_EMAIL }}
          GORGIAS_API_KEY: ${{ secrets.GORGIAS_API_KEY }}
          SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
          SLACK_CHANNEL_ID: ${{ secrets.SLACK_CHANNEL_ID }}