Yes. Productlane classifies incoming threads and can create, prioritize, and route Linear issues without a human touching the case first. No single toggle is called "triage": it's a combination of the Productlane Agent, auto-assignment rules, and the tools you connect. This article lists the pieces and how they fit together.
Two settings let you steer how the AI reads and reacts to incoming cases:
Workspace guidance (Settings > Features > AI > Productlane Agent): tone, policy, and product context the Agent follows when it classifies a conversation and decides whether to act on it.
Custom prompt on AI Drafts and Copilot: shapes how replies and research are drafted once a case reaches an agent. Your custom prompt overrides Productlane's default behavior wherever the two conflict.
See the Productlane Agent article for what workspace guidance controls.
Classification and drafting are grounded in your help center docs, roadmap, and changelog by default. Connecting GitHub adds your repositories as a source. Connecting PostgreSQL lets whitelisted, read-only SQL inform decisions. If none of these sources answer a question, the Agent says so instead of guessing.
Two ways to plug in your own systems and logic:
MCP (Settings > Integrations > API): connects an AI client like Claude or Cursor directly to your workspace, so it can read threads, file issues, and act within the scopes you grant. See the MCP article.
Connectors (Settings > Custom Data): pull data from your own endpoint, like a customer's plan tier or usage, into the thread sidebar so a human or the Copilot has the context to triage without switching tools. See the Connectors article.
AI Copilot acts on connected tools such as Stripe or your database, with your approval, while researching a thread. See the AI Copilot article.
Once a case comes in, these settings decide what happens to it automatically:
Create Linear issues from agent chats, under Productlane Agent settings, files a Linear issue only when the Agent judges the conversation to be a bug or a feature request. It sets a title, description, and priority (Urgent, High, Medium, Low, or No priority), and links the issue back to the thread. See Automatic issue creation.
New issues land in your default Linear team's Triage state (or Unstarted/Todo if Triage isn't available), so your normal Linear workflow picks up from there.
Auto-assign to Linear owner, under Settings > Assignments, routes an incoming thread to the customer's owner in Linear, as long as that owner is a member of your workspace. See Auto-Assignments.
Auto-assign Productlane Agent to bugs hands a classified bug straight to a connected coding agent (Cursor) to open a fix PR. See Code Agents.
If the Agent can't classify or resolve a case confidently, or a customer asks for a person, it hands the thread to your team instead of guessing. The thread stays open, the timeline records the handoff, and your team can be notified over Slack and email. See the "Handoffs to your team" section of the Productlane Agent article.
A typical automated triage setup:
Add workspace guidance describing your product, priority conventions, and any edge cases.
Connect Linear, set a default team, and turn on Create Linear issues from agent chats.
Turn on Auto-assign to Linear owner (Scale plan) so routed issues reach the right teammate.
Connect PostgreSQL, GitHub, or a custom Connector if you want triage decisions grounded in live account or usage data.
Optionally connect Cursor and turn on Auto-assign Productlane Agent to bugs so small, well-scoped bugs get a fix PR without manual assignment.
Each piece is optional and works independently, so you can automate as much or as little of triage as fits your workflow.
