Automation 11/30+: Put an AI agent inside Linear

Description and use

Automation embeds an AI agent in Linear to analyze issues and open pull requests, without altering states, assignments, or merges.

Problem

Linear moves issues quickly but lacks deep understanding of them.

Outcome

The automation provides instant, code‑grounded issue analysis and automates pull request creation, enhancing Linear’s workflow without altering states, assignments, or merges.

Workflow

• Move issue to "analyze" state → AI posts a codebase‑grounded breakdown
• AI comments back, answering in thread and locking scope
• Move issue to "build" state → AI opens a pull request

Evidence

Linear is fast at moving issues. It's still slow at understanding them.

Source

Move an issue into your "analyze" workflow state → the AI posts a codebase-grounded breakdown

Source

Comment back → it answers in the thread and locks down the scope

Source

Move it to your "build" state → it opens a pull request

Source

HMAC-verified webhooks over Linear's GraphQL API; the relay routes every resource type, not just issues

Source

It never changes workflow states, never assigns, never merges.

Source

clone → ./setup.sh → pick Linear → answer a few questions

Source