eq-take-issue

Implements Linear issues by settling the approach with the human before writing code.

Updated Aug 3, 2026
One-click install
npx skills add https://github.com/everquint/frontend-skills --skill eq-take-issue-everquint
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: eq-take-issue
Source: https://github.com/everquint/frontend-skills/tree/main/skills/productivity/eq-take-issue
Command: npx skills add https://github.com/everquint/frontend-skills --skill eq-take-issue-everquint

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Picking up a ticket and jumping straight into code produces implementations that miss the intent, duplicate existing features, or bake in unapproved decisions. This Skill turns a Linear issue into an agreed, test-backed approach before any file is edited, then builds and verifies it against the ticket. ## Core Features & Use Cases - Approach-first workflow: Reads the full issue (comments, blockers, linked docs), maps the codebase seam and prior art, then interviews the human one question at a time until every decision has an explicit answer. - Acceptance criteria as tests: Restates each criterion as a named unit, integration, or E2E test, forming the contract the final verification checks against. - Ticket-native record keeping: Posts the approved approach as an issue comment, claims and branches the ticket, ships a feature doc with the diff, and closes out with PR, issue comment, and deferred-work tickets. - Use Case: Given "implement PROJ-142", the agent reads the issue and its blockers, proposes reusing an existing seam, grills you on data shape and rollout, posts the plan to Linear, then builds it with a failing test first and verifies each criterion before opening the PR. ## Quick Start Ask the agent to pick up and implement Linear issue PROJ-142 using this skill, answering its approach questions one at a time.

Frequently Asked Questions about eq-take-issue

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement a Linear issue with an AI agent?

Give the agent the issue ID or URL. It reads the full issue including comments and blockers, maps the relevant codebase seam, then interviews you one question at a time until the approach is approved. Only after approval does it branch, write failing tests first, and build.

How should acceptance criteria be turned into tests?

Each acceptance criterion is restated as a named test at the unit, integration, or E2E level before coding starts. A criterion with no nameable test is flagged as untestable or not a real criterion and resolved with the human during the approach discussion.

Does this workflow work with any issue tracker besides Linear?

The skill is written specifically around Linear concepts: issue IDs, comments, team statuses, and PR linking that Linear attaches automatically. The approach-first methodology transfers, but the tracker-specific steps assume Linear.

What happens if a discovery breaks the agreed approach mid-implementation?

The discovery goes back to the human as a question rather than into the diff as a unilateral decision. Once resolved, the change of approach is posted to the issue comment so the ticket remains the durable record.

When should I not use a ticket-implementation workflow like this?

Skip it for trivial mechanical changes where the approach interview costs more than the change itself, and for prototypes operating under reduced gates. The workflow is designed for production work where decisions need a durable record.