Linear Integration

The Linear integration connects your AI agent to your Linear workspace, providing both interactive issue management and a fully autonomous development pipeline that takes Linear tickets from creation to pull request.

What You Can Do

Interactive Issue Management

Use the agent to work with Linear issues through natural language:

• Create issues — "Create a bug report for the login timeout issue"
• Search issues — "Find all high-priority bugs assigned to me"
• Update status — "Move issue ABC-123 to In Progress"
• Add comments — "Add a note about the root cause to ABC-123"
• Manage labels — "Add the 'frontend' label to all issues in the Sprint 5 project"
• Link issues — "Make ABC-456 a blocker for ABC-789"

The agent has access to 18 Linear tools covering issues, comments, projects, labels, and relations.

Autonomous Pipeline

The Linear Pipeline is a 10-step automation that takes a Linear ticket from assignment to a reviewed pull request — with no manual intervention required.

How the Pipeline Works

When triggered, the pipeline executes these steps automatically:

Step 1: Triage

The agent reads the ticket and classifies it:

• Type — Bug fix, feature, chore, or refactor
• Complexity — Simple, moderate, or complex
• Requirements — Extracts actionable requirements from the description

Step 2: Preflight Check

Before starting work, the agent verifies:

• The repository is clean (no uncommitted changes)
• The target branch exists and is up to date
• Required tools and dependencies are available

Step 3: Branch Creation

Creates a descriptive branch name from the ticket:

• fix/ABC-123-login-timeout for bugs
• feat/ABC-456-user-dashboard for features

Step 4: Agent Execution

The AI agent runs the two-phase workflow:

1. Planning — Analyzes the codebase and proposes changes
2. Execution — Implements the changes with full tool access

Step 5: Verification

Runs automated checks on the changes:

• TypeScript type checking
• Linting and formatting
• Any configured test suites

Step 6: Self-Review

The agent reviews its own changes:

• Checks for obvious issues
• Verifies the implementation matches the ticket requirements
• Ensures code quality standards are met

Step 7: Pull Request

Creates a GitHub PR with:

• Descriptive title linked to the Linear ticket
• Summary of changes and implementation approach
• References to the original ticket

Step 8: Review Loop

If reviewers request changes:

• The agent reads PR review comments
• Implements requested changes automatically
• Pushes updated commits
• Continues until the PR is approved or a maximum iteration count is reached

Step 9: Slack Notification

Sends a notification to your configured Slack channel with:

• Pipeline status (success/failure)
• Link to the PR
• Summary of what was done

Step 10: Linear Update

Updates the Linear ticket with:

• Final status (completed or needs attention)
• Link to the PR
• Summary comment of the work done

Set Up the Pipeline

Prerequisites

• A Linear account with API access
• A GitHub repository for PR creation
• The desktop app installed and configured

Configuration

1. Open Settings > Connectors in the desktop app
2. Enter your Linear API key
3. Configure the target Linear team and project
4. Set the GitHub repository for PR creation
5. Optionally configure Slack webhook for notifications

Trigger Modes

Timeouts and Limits

Tips

• Start with simple tickets — Test the pipeline with small, well-defined bug fixes before using it for complex features
• Write clear ticket descriptions — The better the ticket description, the better the agent's output
• Monitor initial runs — Watch the first few pipeline runs to calibrate your expectations
• Use labels — Configure which Linear labels trigger the pipeline to control scope