At Linear, agents now create more than half of the issues in its product, up from 3% a year ago, COO Cristina Cordova says. The shift is changing what the company builds and how it thinks about the role of its app: not just a place to store tasks, but a system for coordinating work between people and agents.
Watch Desk analysis
What happened
In an interview with Latent Space, Cordova said Linear’s agent-created issues have climbed past the halfway mark since the company launched its MCP server in May 2025. The company has also built an agent platform and integrations with Codex, Cursor and GitHub Copilot. Cordova says both agent use and use of Linear’s app are rising, rather than one replacing the other.
Her point is that a software product cannot simply sit behind an MCP server and expect to remain useful. If agents can do the work elsewhere, customers still need a reason to return to the product itself to review and coordinate it.
Why it matters
Linear’s example offers a concrete measure of how quickly AI agents are entering everyday software workflows. It also puts a familiar product question in sharper focus: if an agent does the filing, assigning and closing, what is the app for? Cordova’s answer is coordination, but the company’s own usage trends are her evidence, not a guarantee that the model will work for every software business.
The interview also gives a glimpse of the commercial adjustment. Cordova says Linear keeps its agent chat free, charges for some usage-based features, and passes through the published API price plus a small markup for coding sessions using OpenAI or Anthropic models. She says buyers are increasingly choosing one-year deals and paid pilots over longer contracts, while companies create new budgets for AI tools.
Our read
The interesting signal is not that a company has added an agent. It is that agents reportedly went from a sliver to a majority of issue creation in about a year and a half, while the product remains part of the workflow. That is a useful test for software firms: an agent should make the product more valuable, not merely provide a new door into it. Cordova’s account is a strong case study, though the numbers come from the company, not an independent usage audit.
What to watch
- Whether Linear shares more detail on how it measures agent-created work and app usage.
- Whether customers keep returning to coordinate work as agents take on more tasks.
- How one-year AI deals and paid pilots fare when renewal time arrives.
Discussion spark: If agents do more of the work inside a software product, should the company charge for the agent’s usage, the outcomes it delivers, or continued access to the app?
Sources and evidence
- Scaling Linear’s GTM to $100M ARR – Cristina Cordova, COO (19 November 2026, 23:38 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.