What the Omni + Linear integration does
Linear is the hub for engineering work, but an issue rarely contains everything you need to understand it. Context lives everywhere: GitHub, Slack, support tickets, and meeting transcripts.
Omni connects Linear with these sources, allowing an agent to start with an issue and retrieve the related context wherever it lives. Instead of investigating an issue one tool at a time, teams can build a single investigation from the records that actually matter.
How does this work?
The Linear connector brings supported Linear data into Omni. Issues, projects, comments, and other supported records can be retrieved alongside information from other connected systems.
Omni uses hybrid search, combining BM25 lexical search with pgvector semantic search. A search for ENG-142 or OAuthCallbackHandler benefits from exact lexical matching. A broader query such as “users being logged out after token refresh” can surface a GitHub PR describing a “session renewal failure after refresh-token rotation.”
The agent can use both types of results when working through a task. The underlying applications remain the source of truth. Omni handles retrieval and brings the relevant records together.
Omni vs. agents inside Linear
Linear already has its own AI agent, so why bring Omni into the workflow? It comes down to how much information an issue actually contains.
Linear Agent is built around Linear. It understands issues, projects, comments, documents, and workspace history, and can also use MCP to access external tools and data.
Take a production bug. The Linear issue might contain the bug description and current status, but understanding what is actually happening could require information from several places:
- Linear: issue details, project status, dependencies
- GitHub: related PRs and code changes
- Support: customer reports and affected workflows
- Slack: investigation notes, workarounds, and ongoing discussions
- Meetings: technical decisions and the reasoning behind them
Omni brings these sources together around the issue. The agent can trace related records across connected systems and use them to build a fuller picture of the problem, its history, and its current state.
The result is more than a summary of the Linear issue. It is a connected view of the information surrounding it.
What can an AI agent do with Linear data?
The useful workflows are the ones where the Linear issue is only the starting point.
Release status: Linear + GitHub
A Linear project shows the work planned for a release. GitHub shows what is actually happening with the code.
Connecting the two makes it possible to identify:
- Release issues without an active PR
- PRs still waiting for review
- Issues marked In Progress even though their PR has already merged
- Dependencies holding up other work
- Work that is ready, blocked, or still needs attention
This gives the agent enough information to distinguish between what is planned in Linear and what has actually moved forward in the codebase.
Bug triage: Linear + support
A Linear issue describes the engineering problem. Support conversations show how customers are experiencing it.
Suppose several customers report checkout failures that appear unrelated in Linear. Looking across support tickets could reveal that they all occur during the same payment flow.
With both sources connected, the agent can link related customer reports to existing Linear issues, group similar symptoms, identify affected workflows, and bring the customer evidence into the engineering investigation.
What did we decide? Linear + meeting transcripts
The final decision may be recorded in Linear. The reasoning behind it may still live in a meeting.
An authentication issue, for example, might document the architecture the team chose without explaining the trade-offs behind it. The meeting transcript could contain the discussion around latency, backward compatibility, security requirements, or infrastructure constraints.
Connect Linear to meeting transcripts, and the agent instantly traces a decision back.
How to connect an AI agent to Linear
You can connect an AI agent to Linear using a Personal API key. Linear lets you control the permissions and, where applicable, team access associated with the key.
In Linear
- Create a Personal API key
Go to Settings → Account → Security & Access and create a new API key. - Label the key
Use a name such as Omni so it is easy to identify later. - Set permissions and team access
Give the key only the read permissions and team access required for the workflow. - Copy the key securely
Use it when setting up the Linear connection in Omni.
In Omni
- Open the Linear connector and enter the requested Linear credentials.
- Once the connection is established, Omni can synchronize the authorized Linear data and retrieve it alongside information from other connected systems.
- When you connect an AI agent to Linear, the API key defines what the connection can access. Omni then makes that authorized data available alongside the rest of the connected stack.
Permissions: What Omni reads from Linear
Connecting Linear to Omni does not mean turning the entire workspace into an unrestricted data pool.
The connector works with the permissions granted through the Linear connection, while Omni’s architecture carries source permissions into its indexed data.
This matters when Linear is connected alongside systems such as GitHub, Slack, or Notion. Each system has its own access rules, and those boundaries still need to hold when information is brought into a shared retrieval layer.
Omni’s connector architecture exposes the permission pipeline rather than hiding it behind a closed service. Connectors can emit document ACLs and group-membership changes through the connector SDK.
That makes the implementation available for inspection and audit. Omni’s core is open source under Apache 2.0. See the Omni open-source project to inspect the implementation and connector architecture.
Works better together
A Linear issue rarely exists in isolation. The code may be in GitHub, the customer context in support, and the reasoning behind a decision in a meeting transcript.
Omni connects these sources so an agent integration can follow the context around a Linear issue instead of stopping at the issue itself.
For example:
- Linear + GitHub: GitHub connector connects an issue to its PRs, commits, and implementation status.
- Linear + Fireflies: Fireflies connector traces an issue back to the meeting where an architectural or product decision was made.
- Linear + ClickUp: ClickUp connector connects engineering issues with customer reports and affected workflows.
The agent can work from the Linear issue while pulling in the GitHub, meeting, or support context needed to understand it.
Frequently asked questions
Can Omni explain why a release is blocked using Linear issues and GitHub PRs?
Yes, when Linear and GitHub are both connected and the relevant content has been indexed. Omni can retrieve issue status, comments, and project updates alongside related pull requests and discussions from other connected sources. It helps assemble the evidence behind a blocker; it does not change issue status or merge a pull request through the Linear connector.
What permissions should I grant the Linear API key?
Use a Linear API key with read access for the content you want to index. The connector only retrieves data and does not need write permissions. Treat the key like a password, because it grants access to workspace content. Optional team keys can narrow issue fetching, but do not restrict the workspace-wide fetching of projects and documents.
Does team filtering restrict everything the Linear connector syncs?
No. Team keys limit issue syncing, while projects and documents are fetched workspace-wide. Linear permission metadata is retained where available, but projects and documents without retained team keys are indexed as public within Omni. Do not treat the issue filter as a complete security boundary for all Linear content; review the data and permission behavior before connecting a sensitive workspace.
What does Omni add if we already use Linear Agent?
Omni adds context from outside Linear. An employee can investigate an issue together with related Slack conversations, customer records, meeting transcripts, and company documents rather than asking only about the engineering workspace. Linear Agent and Omni serve different scopes; connecting Linear to Omni does not replace Linear’s own agent or give this connector write access.
Which Linear content can Omni index?
Omni indexes issues, projects, project updates, and documents. Issue content includes titles, descriptions, statuses, priorities, assignees, labels, and comments. The API key’s access and any configured issue team filters determine what the connector can fetch; content must be synced before it can appear in Omni’s index.
Can Omni create or update Linear issues?
Not through this connector. The Linear integration is read-only and cannot create, modify, or delete issues, projects, or documents. It is designed for retrieval and cross-tool answers rather than issue-management automation.
What do I need to connect Linear to self-hosted Omni?
Run the Linear connector service with your Omni deployment, create a read-access Linear API key, and enter it under Settings → Integrations → Linear. Optionally provide team keys to limit issue syncing, then start the initial sync. Review the separate project and document permission behavior before enabling the connection.