issue-triage

Automate Linear issue triage by running the /issue-triage command with desired flags to analyze your backlog. It identifies duplicates, flags stale issues, and validates priorities using semantic similarity and multi-agent consensus, outputting a Markdown triage report and JSON audit log.

Detect duplicate Linear issues using semantic similarity to surface near-duplicates across your backlog. The triage process scans large backlogs across teams like ENG and PRODUCT to identify matching issues and outputs an actionable prioritized triage report with confidence scores.

Validate Linear issue priorities automatically using an impact and urgency rubric with multi-agent consensus. The triage process applies parallel agent integration for ambiguous decisions to verify priority alignment across large ENG and PRODUCT backlogs and surfaces actions with confidence scores.

Automated Linear backlog triage requires Linear MCP configuration to access your issues. It also uses Python tooling and jq for processing, generating a Markdown triage report and a JSON audit log as outputs to keep work well-scoped and actionable.

Staleness detection for Linear issues works by checking for inactivity and missing file references within the backlog. The automated triage flags these stale issues so teams can surface necessary actions and maintain a well-scoped backlog.