Skill

yux-pm-triage

Triage Linear inbox issues with AI classification. Triggers on "pm triage", "triage inbox", "classify issues", "process inbox", "处理反馈".

npx claudepluginhub wuyuxiangx/yux-claude-hub --plugin yux-linear

Tool Access

This skill is limited to using the following tools:

ReadWriteGlobGrepBash(git:*)Bash(gh:*)mcp__linear__*

Preview

Process inbox issues: classify type, set priority/effort, structure descriptions, and move to Backlog.

SKILL.md

Similar Skills

auto-triage

Classifies, deduplicates, prioritizes, and dispatches raw incoming issues into actionable work items for dev agents. Use for GitHub issues or external sources.

5 files

zero-review

triage

64.1k

Triages project issues via label state machine with bug/enhancement categories and states like needs-triage, needs-info, ready-for-agent. Use for reviewing bugs, features, or workflows.

2 files

mattpocock-skills

linear

Manages Linear issues via linearis CLI: list, search, read details, create with options, update title/description/status/labels/assignee. Integrates with dev workflows.

johnlindquist-claude

Stats

Parent Repo Stars0

Parent Repo Forks0

Last CommitMar 30, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

PM Triage

Process inbox issues: classify type, set priority/effort, structure descriptions, and move to Backlog.

Prerequisite: .claude/linear-config.json must exist. If missing, show: "Please run /yux-linear-init first."

Workflow

Step 1: Load Config and Labels

Read .claude/linear-config.json

mcp__linear__list_issue_labels(team: "<team.id>")

Parse labels into categories:

Type labels: Feature, Bug, Epic, Chore, Improvement
Func labels: Frontend, Backend, API, Extension, ML

Step 2: Fetch Inbox

mcp__linear__list_issues(
  project: "<project.id>",
  state: "Triage",
  limit: 20,
  orderBy: "createdAt"
)

If empty, report clean inbox and stop.

Display inbox summary table (ID, title, created date) and ask user to process all or select specific items.

Step 3: AI Analysis

For each selected issue, analyze these factors:

Type Detection (keyword-based):

Bug: "crash", "error", "fail", "broken", "崩溃", "错误", "失败"
Feature: "add", "new", "support", "希望", "能不能", "新增"
Improvement: "improve", "better", "optimize", "优化", "改进"
Chore: "update", "upgrade", "migrate", "更新", "升级"

If multiple types match, prefer: Bug > Feature > Improvement > Chore.

Priority Assessment:

Urgent: Security issues, data loss, complete feature broken
High: Core flow affected, many users impacted
Medium: Feature request with clear value, non-critical bugs
Low: Nice-to-have, cosmetic issues

Effort Estimation (T-shirt sizing):

XS: < 2 hours (typo fix, config change)
S: 2-8 hours (small bug, minor feature)
M: 1-3 days (medium feature, complex bug)
L: 3-7 days (large feature, refactoring)
XL: > 1 week (epic-level work)

Step 4: Confirm Per Issue

For each issue, display:

Detected type, priority, effort, labels
Structured description (Problem / Context / Acceptance Criteria)

Ask user to apply, reject, or edit before proceeding.

Step 5: Apply Changes

On confirmation, update the issue in Linear:

mcp__linear__update_issue(
  id: "<issue_id>",
  description: "<structured_description>",
  priority: <priority_number>,
  labels: ["<type_label_id>", "<func_label_id>"],
  project: "<project.id>",
  state: "Backlog"
)

Step 6: Summary

After all items are processed, display summary:

=== Triage Complete ===

Processed: 5 issues

By Type:    Bug: 2  |  Feature: 2  |  Improvement: 1
By Priority: High: 1  |  Medium: 3  |  Low: 1

Moved to Backlog:
  WYX-120  [Bug/High]    Login crash on iOS
  WYX-121  [Feature/Med] Dark mode support
  WYX-122  [Bug/Med]     Typo in settings