Linear ships fast and integrates broadly, which is a good problem until you open the integrations directory and find roughly a dozen categories staring back at you. Most teams respond by turning on the four they recognize and ignoring the rest, which is not a terrible strategy. It just leaves value on the table in one or two specific places, and it hides a gap that matters if your customer problems arrive somewhere other than Linear.
This guide is the practical version. It covers how the directory is actually organized, which Linear integrations are worth setting up on day one, which ones to wait on, and what to do about the category with the most obvious hole in it.
In this article
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
What Counts as a Linear Integration
The word is doing a lot of work. Two things in the same directory can mean completely different levels of connection, and the difference decides whether the integration is load bearing or decorative.
A notification wire tells a human that something happened. Linear's Slack integration is the archetype: an event fires in Linear, a formatted message lands in a channel, and a person reacts. Nothing is stored on the other side. If the integration breaks, you lose awareness, not data.
A record sync keeps two systems holding a version of the same thing in step. Linear's GitHub integration is the archetype here: a branch name maps to an issue, the pull request state drives the issue state, and both sides stay accurate without anyone updating them by hand. If this breaks, two systems disagree about reality and someone acts on the wrong one.
How the Linear Integrations Directory Is Organized
The categories overlap on purpose, and several tools appear in more than one. Here is what each is really for.
| Category | What it does | Representative tools |
|---|---|---|
| Essentials | The connections most teams turn on first | GitHub, Slack, GitLab, Figma, Intercom, Google Sheets |
| Engineering | Code, errors, and editor surfaces | GitHub, GitLab, VS Code, Sentry, Datadog, Raycast |
| Customer Experience | Support and revenue tools feeding Linear | Zendesk, Intercom, Front, Salesforce, Attio, Gong |
| Bug Reporting | Capturing a defect with enough context to act | Linear Asks, Sentry, incident.io, Jam, Bird Eats Bug |
| Automations | Rules and bulk movement of issues | Zapier, issues via email, Jira, Fivetran |
| Analytics | Getting Linear data somewhere you can query it | Airbyte, Google Sheets, Fivetran, Retool, Jellyfish |
| Collaboration | Docs, chat, and search | Slack, Notion, Discord, Linear Asks, Glean |
| Agents | Coding tools that pick up an issue and work it | Codex, Cursor, GitHub Copilot, Devin, Factory |
| AI clients | Assistants that query Linear over MCP | Claude, ChatGPT, Cursor, Windsurf, Replit |
| Media and Design | Visual context attached to issues | Figma, Loom, Miro, Descript, YouTube |
| Security and Compliance | Evidence collection for audits | Vanta, Drata, Fencer, SecureSlate |
Two things stand out. First, the engineering half of the directory is mature and mostly first party. Second, the customer-facing half is thinner than it looks, and it has a specific hole in it that we will get to.
The Linear Integrations Almost Every Team Should Turn On
Four connections carry most of the value. If you set up only these, you have a working Linear.
GitHub or GitLab
The highest-leverage integration Linear offers. Branch names and pull requests link to issues automatically, and merging moves the issue without anyone touching it. This is real record sync, not a notification, and it is the reason Linear's status column stays honest on teams that use it.
Slack
The notification surface. Scoped subscriptions push the events a channel can act on, and the slash command and message shortcuts let anyone file an issue from a thread. The capture half is where it earns its keep and the half most teams never use. The Linear Slack integration guide covers the setup that avoids channel noise.
Figma
Cheap to set up and immediately useful. Design files embed in issues with live previews, so the engineer working a ticket sees the current design rather than a screenshot someone pasted three weeks ago.
Sentry or your error reporter
Turns a production error into a Linear issue with the stack trace attached, and links the issue back so a resolved error closes the loop. It is the difference between finding out about a crash from a customer and finding out from your own tooling.
Everything past these four should earn its place. The failure mode is not a missing integration. It is six integrations firing events that nobody reads, which trains the team to ignore the ones that matter.
Linear Integrations for Customer Support
This is the category worth reading carefully, because it is where Linear's model differs from a normal issue tracker.
Linear does not want your support tickets. It wants the customer request: a lightweight record that attaches a customer, and their revenue and tier, to an issue, so the backlog can be ranked by who is actually affected rather than by who complained loudest. Zendesk, Intercom, and Front feed that primitive from the support side. Salesforce, Attio, and Gong feed it from the revenue side. The Linear Customer Requests guide covers how the primitive works, and Linear Triage covers the queue everything lands in.
That design is deliberate and it is good. It also means Linear is not trying to be your help desk, which is the honest answer to a question a lot of teams ask before they get here. Linear as a ticketing system walks through what it genuinely handles and the four things it does not.
The Agent and AI Client Categories
These are the newest additions and the ones most likely to change shape between now and next year. They split cleanly.
Agents are coding tools that take a Linear issue as their unit of work. Codex, Cursor, GitHub Copilot, Devin, Factory, and Sentry Agent all sit here. You assign an issue to an agent the way you would assign it to a person, and it opens a pull request.
AI clients connect an assistant to your Linear workspace over MCP so you can ask questions and make changes in conversation. Claude, ChatGPT, Cursor, Windsurf, Replit, and Vercel's v0 connector are the listed options.
The useful posture on both is assistive rather than autonomous. An agent that drafts a fix for a well-specified bug saves real time. An agent working a vague issue produces a pull request someone has to review and usually discard, which costs more than it saved. The quality of the issue decides the quality of the output, which is a familiar rule with a new surface.
An agent is only as good as the issue it picks up. Integrations that improve issue quality are worth more than integrations that add throughput.
Automations, Analytics, and the Long Tail
Three groups here, with different reasons to use them.
Automations cover Zapier, creating issues from an email address, and the Jira connection used mostly for migration. Zapier is the one to be careful with. It is excellent at a one-shot trigger, such as creating an issue when a form is submitted, and poor at keeping two records in step over the life of a bug. That job needs conflict resolution and loop prevention, and a chain of Zaps has neither. Bidirectional sync explains what that actually requires and how to tell a real two-way sync from a pair of one-way rules wearing a costume.
Analytics covers Airbyte, Fivetran, Google Sheets, Retool, and Jellyfish. Skip these until someone asks a question about cycle time or throughput that Linear's own insights cannot answer. When that day comes, a warehouse pipeline is the right answer and a spreadsheet export is not.
The long tail is Security and Compliance, Media and Design, and the assorted partner tools. Vanta and Drata are worth it the week you start a SOC 2 audit and not before. Loom and Miro are worth it if your team already lives in them.
Which Linear Integrations Are Worth Your Time
The honest answer depends on where your work comes from, not on how many tools you run.
GitHub or GitLab for real status sync, Slack for notifications and capture, Figma for design context, and an error reporter like Sentry. Four integrations, all first party, all low maintenance. This covers the engineering loop end to end.
A support connection when engineering keeps asking who reported a bug. A warehouse pipeline when someone asks about cycle time twice. Notion when specs keep getting pasted into issue descriptions. Wait for the second occurrence, then wire it.
Compliance tools before an audit is scheduled, analytics before anyone asks a question, and most of the partner long tail. These are easy to turn on and easy to forget, which is exactly how a workspace ends up with integrations nobody can explain.
What to Do When Linear Has No Native Integration
Which brings us back to the hole in the Customer Experience category.
If support runs in HubSpot Service Hub and engineering runs in Linear, you are connecting two tools that do not know about each other. There are three ways to do it, and they are not equivalent.
An automation builder like Zapier or Make can create a Linear issue when a HubSpot ticket is created. That covers the first hop and nothing after it. Status changes, comment threads, and the customer-facing update all need their own rules, and a minimum viable two-way setup is three or more automations your team now owns and maintains.
A general sync platform like Unito does mirror records both ways. It is the right call when HubSpot to Linear is one of several syncs you run. For a single flow, you are configuring field maps and paying platform-tier pricing for one connection.
A purpose-built app inherits the workflow instead of assembling it. IssueLinker creates a Linear issue from a HubSpot ticket in one click, syncs status and comments both ways with loop prevention, and keeps the customer-facing reply ready for the moment the fix ships. It works without HubSpot Pro+ workflows, which matters more than it sounds like it should, because that is where most native-workflow approaches quietly stop.
The full comparison lives in the best HubSpot to Linear integrations, and the four approaches guide walks each one end to end.
Linear does not integrate with HubSpot. IssueLinker does.
If support runs in HubSpot Service Hub and engineering runs in Linear, IssueLinker fills the gap Linear's directory leaves open. One-click create and link from the ticket, two-way status and comment sync, and the customer hears about the fix the moment it ships.
A Five-Minute Integration Audit
If Linear has been running for a while and the integration list has grown without anyone pruning it, three quick checks usually surface the problem.
- 1
The decision audit
List every integration currently enabled and write the decision each one changes next to it. Any integration with a blank next to it is a candidate for removal. Turning it off costs nothing, and you will know within a week if you were wrong.
- 2
The status audit
Open ten recently closed issues and check whether the status moved on its own or a human moved it. If humans are doing the moving, your source control integration is either not set up or not being used, and that is the highest-value thing on this page to fix.
- 3
The origin audit
Take the last twenty issues and ask where each one came from. If a meaningful share arrived by someone copying a customer problem out of another tool by hand, the missing integration is on the customer side, not the engineering side.
That third check is the one that surprises teams. The engineering integrations get set up on day one because engineers set them up. The customer-facing connection is the one nobody owns, and it is usually the one costing the most time.
