linear

Overview of linear skill

The linear skill helps you work with Linear through the Linear MCP server, so an agent can read, create, and update issues without guessing at workspace structure or workflow conventions. It is most useful for people who need linear for Issue Tracking, triage, project coordination, or keeping team work aligned to live issue state.

What this skill is for

Use this linear skill when the task depends on accurate Linear data: finding the right issue, updating status, adding context, assigning ownership, or checking project progress. It is better than a generic prompt when you need the agent to operate inside a real Linear workspace instead of drafting a stand-alone recommendation.

Who should install it

Install linear if you already use Linear for issue tracking and want an agent to help with operational work: triage queues, ticket hygiene, project follow-up, or coordination across teams. It is a strong fit for users who care about workflow correctness, not just message generation.

Main decision factors

The key dependency is access: the skill requires a connected Linear MCP server and OAuth login. If your workspace, teams, or projects are restricted, adoption can stall until permissions are confirmed. The biggest value is reliable, context-aware Linear usage; the biggest risk is assuming the skill can act before MCP is configured.

How to Use linear skill

Install and connect Linear MCP

For linear install, add the skill and connect the MCP server before expecting results:

codex mcp add linear --url https://mcp.linear.app/mcp
codex --enable rmcp_client
codex mcp login linear

If login succeeds, restart Codex before continuing. On Windows connection issues, the WSL-based MCP transport may be the faster path to a working setup.

Give the skill the right task shape

The best linear usage starts with a concrete action, not a vague goal. Good inputs name the object and the desired change: “Find the bug report for login failures, move it to In Progress, and summarize blockers,” or “Create a Linear issue for onboarding docs with owner, priority, and acceptance criteria.” Avoid prompts like “organize Linear” unless you also specify scope, team, and what output you want.

Read these files first

Start with SKILL.md, then check agents/openai.yaml for the default intent and MCP dependency. Also review LICENSE.txt for reuse context and the asset files only if you need branding. The main operational detail is the required workflow in SKILL.md, especially setup and clarification steps.

Workflow tips that improve results

When you want the model to act in Linear, provide the workspace, team, project, and whether the request is read-only or mutating. If you need issue updates, include the exact field changes you want, plus any constraints like priority, due date, or assignee. If the MCP connection is not ready, the skill should stop and help you finish setup before continuing.

linear skill FAQ

Is linear only for issue tracking?

No. linear for Issue Tracking is the common case, but the skill also fits project updates, team coordination, and documentation-linked workflow tasks inside Linear. Use it whenever the source of truth is in Linear and the output needs to reflect live workspace state.

What is the difference between linear skill and a normal prompt?

A normal prompt can describe what should happen; the linear skill is built to do it against a connected Linear workspace. That matters when you need real issue IDs, current status, permissions, or reliable updates rather than a text-only plan.

Is this beginner-friendly?

Yes, if you can name the task clearly and you already have access to Linear. It is less beginner-friendly if you are still deciding which team, project, or issue should be changed, because the best linear guide inputs are specific.

When should I not use it?

Do not use linear if you only need a generic project-management template, if your Linear workspace is not connected, or if the task should stay outside your issue tracker. It is also not the right fit when you need broad ideation without any live Linear action.

How to Improve linear skill

Provide richer Linear context

The strongest linear guide inputs include the workspace, team, project, issue title or link, and the exact operation. If you want a better triage result, include symptoms, priority hints, and any related issue references so the agent can disambiguate quickly.

State the desired end state

Instead of asking for “help with this issue,” say what the Linear issue should look like after the task: status changed, owner set, description rewritten, acceptance criteria added, or duplicates merged. Clear end-state instructions reduce back-and-forth and make Linear usage more predictable.

Watch for setup and permission failures

The most common failure mode is not the prompt itself but missing MCP access or insufficient workspace permissions. If the skill cannot connect, fix the Linear MCP setup first; if it can connect but cannot act, verify the relevant team and project permissions before retrying.

Iterate with tighter prompts after the first pass

If the first result is too broad, narrow it by adding one issue, one team, or one project at a time. For repeated work, reuse a stable prompt pattern that includes task type, scope, and success criteria so the linear skill can produce consistent updates with less editing.