Lindesk

A CLI tool to integrate Zendesk tickets with Linear issues using AI summarization.

Features

• Fetch ticket details from Zendesk
• Analyze complete ticket conversations using Amp AI
• Generate detailed, well-structured technical summaries for engineers
• Create a Linear issue with the AI-generated summary
• Link back to the original Zendesk ticket

Prerequisites

• Node.js 16+
• Access to Zendesk (admin privileges to generate API tokens)
• Linear workspace access (ability to create API keys)
• Amp account with API access

Installation

npm install -g lindesk

Or run directly:

npx lindesk

Setup

Before using Lindesk, you need to configure your API credentials:

lindesk setup

This will prompt you for:

• Zendesk domain, email and API token
• Linear API key
• Amp API key and endpoint
• Optional default Linear project

Alternatively, you can use environment variables:

ZENDESK_DOMAIN=your-company.zendesk.com
ZENDESK_EMAIL=your-email@example.com
ZENDESK_TOKEN=your-zendesk-token
LINEAR_API_KEY=your-linear-api-key
AMP_API_KEY=your-amp-api-key
# AMP_ENDPOINT is optional, uses Amp CLI by default
DEFAULT_LINEAR_PROJECT=TEAM  # Optional

Usage

Transfer a Zendesk ticket to Linear

lindesk transfer <ticketId> --project <linearProjectKey>

Example:

lindesk transfer 12345 --project ENG

If you've set a default project during setup, you can omit the --project flag:

lindesk transfer 12345

How it works

1. Fetches the complete ticket details and conversation history from Zendesk
2. Analyzes the entire conversation using Amp AI to create a comprehensive technical analysis
3. Generates a detailed, well-structured technical summary including:
   • Problem summary and environment details
   • Step-by-step reproduction instructions
   • Investigation findings and impact assessment
   • Priority and complexity assessment
4. Creates a new Linear issue with the AI-generated analysis and sets appropriate priority
5. Links back to the original Zendesk ticket for reference

Output Format

Lindesk creates Linear tickets with structured, detailed descriptions that include:

• Problem Summary: Clear overview of the issue
• Environment: Technical environment details
• Reproduction Steps: Numbered, step-by-step instructions
• Expected vs Actual Behavior: What should happen vs what actually happens
• Investigation Findings: Technical details discovered during analysis
• Impact: Assessment of user and business impact

License

MIT