Surfaced reads Linear tickets before they reach Codex, Cursor, Claude Code, Linear Agent, or your internal AI workflow, then surfaces the product and engineering decisions that would otherwise become rework, review cycles, and human cleanup.
Surfaced reads the ticket before an AI coding tool does, and shows the missing decisions before they come back as review work.
Every run returns a product framing and an implementation framing. Product sees the assumptions about user needs and acceptance criteria. Engineering sees the technical judgment calls and missing constraints.
Each flag is a question someone can resolve with a short reply. The answer stays attached to the ticket, so the context lives where the work happens.
Surfaced listens for Linear activity and posts its analysis as a comment on the ticket. The result is a clearer handoff whether the next step is Codex, Cursor, Claude Code, Linear Agent or an internal agent workflow.
See the estimated ticket count, character volume, and a per-model cost range before you confirm. You decide what runs. Estimates are rough during beta and improve as we calibrate.
Connect your own key for Claude Sonnet, OpenAI GPT, Gemini Pro, DeepSeek or Qwen. Keys are encrypted at rest and used only for the analyses you set up, including the ones that run on a webhook when you are away.
Track assumptions resolved and an estimated range of staff time saved. No per-person scores, no leaderboards, no ranking your team against each other.
Install Surfaced in Linear and add your AI provider key. During a short onboarding you set the analysis scope and your baseline inputs for the time-saved estimate.
When a ticket is headed toward AI implementation, Surfaced surfaces the assumptions and judgment calls in the spec, framed once for product and once for engineering.
Flagged assumptions turn into questions. Product and engineering answer the ones that change the work, right inside Linear.
Tickets move into Codex, Cursor, Claude Code, Linear Agent coding sessions or your own workflow with fewer open questions. Your dashboard shows assumptions resolved and an estimated range of time saved.
Your first week is a baseline, not a grade
Week one surfaces assumptions that were always in your tickets. That is your starting point. Many teams see even more in week two as more detail gets written down, and that is the tool working, not your team slipping.
Surfaced is free to use while we are in beta, and because you bring your own AI key, model usage is billed by your provider, not by us. The plans below are what we are planning for after beta. They are not final, and we would like your feedback.
Indicative only. Nothing here is billed during beta.
Free
Individual Pro
Team
Scale
Enterprise
Lifetime memberships for early supporters
Beta users will get the chance to buy a lifetime membership before public launch. A limited number of lifetime memberships will also be offered to early post-beta users to help support the launch. The Individual Pro lifetime membership is planned at $79, and the Team lifetime membership at $250.
Euro prices are roughly 10% below the US dollar prices. Need more analyses on any plan? Top-ups will be available for any amount, with pricing to be confirmed.
Every plan uses your own AI provider key, so you only ever pay your provider for model usage. Have a view on any of this? Send us your feedback.
Install Surfaced in Linear and catch the missing decisions before they become review cycles.
Free during beta · Bring your own AI key · Ready for your agent workflow