An engineer files a bug in Slack. Someone converts it to a Linear issue. The issue lands in triage with a title, a one-line description, and nothing else - no assignee, no label, no project, no priority. Before anyone can act on it, someone has to read it, think about it, and fill in the blanks. At volume, that's not a speed bump. It's a recurring tax on whoever is running triage that week.
Linear's design goal has always been to eliminate the friction that plagues most project management tools. It's fast, keyboard-first, and opinionated about workflow. But the speed of the interface doesn't fix the problem of incoming work that arrives without context. That problem lives upstream, in the moment when someone frames a Slack message and a human (or a one-click action) converts it into a ticket.
Where the metadata gap comes from
When your development work happens in Linear and your team's conversations live in Slack, you create a gap.
The native integration has three useful mechanisms:
the /linear slash command creates an issue from inside a thread; message shortcuts let you turn any Slack message into an issue in two clicks; and URL unfurling renders Linear issue links inline with status, assignee, and the latest update.
The shortcut path is fast. That's part of the problem. Two clicks is not enough time to pick a project, assign a team, or set a priority. You get an issue with a title. The rest is left for triage.
If your product development process requires marketing, customer success, or operations to participate in issue tracking, those teams will struggle with Linear's vocabulary and mental model. The issue they file won't have the right label. They may not know which team owns the area. The priority will be blank, or wrong.
Without triage, unvetted issues can clutter sprint backlogs and slow down team velocity. That's the reason the triage queue exists. But if triage is where metadata gets added from scratch on every ticket, the queue becomes a second job.
What Triage Intelligence actually does
Linear shipped Triage Intelligence to address this directly. When enabled, issues in your workspace are analyzed by agentic models, and every future issue that enters triage is assessed against the rest of your data, which allows it to proactively surface suggested issue properties and relationships.
The suggestions are specific: issue properties suggested include teams, projects, assignees, and labels, and can be configured to auto-apply when suggested.
The engineering behind it is worth understanding. It uses a combination of search, ranking, and LLM-based reasoning to make suggestions as new issues come in, drawing on your existing backlog as a dataset to understand how similar work has been organized in the past. Linear noted that they switched to larger models like GPT-5 and Gemini 2.5 Pro, which could handle more complexity, unlocking better suggestion quality - especially for fuzzier or more complex issues.
Beyond detecting duplicates, Triage Intelligence can now reliably suggest related issues, recommend properties like labels or assignees, and explain its reasoning.
How Linear uses it internally: after the automated pass, the person overseeing triage makes the final decision about where each issue should go - internally called "the goalie." Engineers and product team members rotate through the role each week, and for bugs, the goalie either takes the issue themselves or assigns it to the right expert.
The part the tool can't do on its own
Triage Intelligence handles the routing problem once an issue exists. It doesn't help with the moment before - when a Slack thread contains enough information to write a well-formed issue, but nobody has done it yet.
Linear supports cross-posting initiative updates to Slack, and comments and reactions sync bi-directionally between both tools. That's useful for keeping non-Linear users informed. But the reverse problem - converting Slack conversations into well-scoped issues proactively - still requires a human to notice and act.
This is where an AI teammate in Slack can quietly close the loop. A tool like Beagle sits inside the conversation and can watch for threads that look like bug reports or feature requests - threads that haven't been tracked anywhere yet - and prompt the right person to file them with context already drafted. The issue title, a one-paragraph description, and a suggested project, pulled from the thread. The human still decides whether to send it. But the blank-field problem is gone before triage ever sees it.
When an AI-enriched issue is filed, it can generate a short summary of the thread, draft a title from the conversation, and carry over the original messages, attachments, and requester details instead of making someone copy-paste everything by hand.
The goalie problem at scale
As organizations scale, questions like "who should work on this" or "what labels are usually applied in these scenarios" become harder to answer. The triage rotation helps distribute the cognitive load. But it also means institutional knowledge about routing is re-learned by whoever is on duty that week.
You can automate much of the process of sorting incoming work with triage rules, which let you define actions that run the moment after Triage Intelligence completes its work - for example, routing all urgent enterprise bugs directly to a specific team or assignee and automatically applying the right SLA.
That's the right shape: AI makes the suggestion, rules encode the team's accepted conventions, humans handle the edge cases. The metadata gap doesn't disappear, but it shrinks to the issues that are genuinely ambiguous. Which is a much better use of the goalie's time.