issue analysis with label inference

🏷️ Issue Analysis with Label Inference

Version: 2.0.0 Last Updated: 2025-11-22 Priority: ⭐⭐⭐⭐ (P1 Level) Purpose: AI-driven Issue analysis and automatic inference from the 57-label system

📋 Overview

Executes AI-driven Issue analysis and automatic label inference based on Miyabi's 57-label system (11 categories).

🎯 P0: Invocation triggers

Trigger	Example
Label question	"what labels should I use?"
Issue analysis	"analyze this Issue", "triage issue #270"
Label inference	"infer labels for this Issue"
After Issue creation	Automatic processing after Issue creation
Backlog management	"prioritize backlog"

🔧 P1: Label category list

11 categories, 57 labels

#	Category	Count	Required/Optional	Purpose
1	STATE	8	Automatic	Lifecycle management
2	AGENT	6	Recommended	Agent assignment
3	PRIORITY	4	Required	Priority management
4	TYPE	7	Required	Issue classification
5	SEVERITY	4	Conditional	Severity (bugs only)
6	PHASE	5	Recommended	Project phase
7	SPECIAL	7	Conditional	Special flags
8	TRIGGER	4	Automatic	Automation triggers
9	QUALITY	4	Automatic	Quality score
10	COMMUNITY	4	Recommended	Community
11	HIERARCHY	4	Automatic	Issue hierarchy

🚀 P2: Label inference rules

TYPE inference (required, 1)

Keyword	Label	Example
"add", "implement", "create", "new"	`✨ type:feature`	New feature addition
"fix", "crash", "error", "broken"	`🐛 type:bug`	Bug fix
"docs", "README", "guide"	`📚 type:docs`	Documentation
"refactor", "cleanup", "reorganize"	`🔧 type:refactor`	Refactor
"test", "coverage", "e2e"	`🧪 type:test`	Test
"architecture", "system design"	`🏗️ type:architecture`	Architecture
"deploy", "CI/CD", "docker"	`🚀 type:deployment`	Deploy

PRIORITY inference (required, 1)

Condition	Label	SLA
Security, production outage, data loss	`🔥 priority:P0-Critical`	24 hours
Major feature, critical bug, performance degradation	`⚠️ priority:P1-High`	3 days
Normal feature, standard bug	`📊 priority:P2-Medium`	1 week
Minor improvement, typo	`📝 priority:P3-Low`	None

SEVERITY inference (conditional, bugs only)

Condition	Label	Escalation
Production halt, security breach	`🚨 severity:Sev.1-Critical`	Guardian + CISO
Major feature failure	`⚠️ severity:Sev.2-High`	TechLead
Partial feature issue	`📊 severity:Sev.3-Medium`	Agent automatic
Minor UI issue	`📝 severity:Sev.4-Low`	Agent automatic

SPECIAL inference (conditional)

Keyword	Label	Action
"XSS", "SQL injection", "CVE"	`🔐 security`	CISO notification, Issue private
"Claude API", "high cost"	`💰 cost-watch`	Budget monitoring
"depends on #", "blocked by"	`🔄 dependencies`	Wait until dependency resolved
"research", "investigate"	`🎓 learning`	SLA extension
"experiment", "POC"	`🔬 experiment`	Failure allowed

COMMUNITY inference (recommended)

Condition	Label
Within 2 hours, no dependencies, clear requirements	`👋 good-first-issue`
External expertise required	`🙏 help-wanted`

HIERARCHY inference (automatic)

Condition	Label
No parent Issue	`🌳 hierarchy:root`
Has child Issue	`📂 hierarchy:parent`
Has parent Issue	`📄 hierarchy:child`
No child Issue (terminal)	`🍃 hierarchy:leaf`

⚡ P3: Analysis workflow

Step 1: Read Issue content

- タイトル
- 本文（説明）
- コメント（あれば）
- 関連Issue（リンク）

Step 2: Extract keywords

- 技術キーワード: "security", "performance", "API"
- アクションキーワード: "add", "fix", "refactor"
- 緊急度キーワード: "urgent", "critical", "nice-to-have"

Step 3: Apply inference rules

1. TYPE決定（必須、1個）
2. PRIORITY決定（必須、1個）
3. SEVERITY決定（バグ/インシデント時）
4. SPECIAL条件チェック
5. COMMUNITY適合性評価
6. HIERARCHY位置決定

Step 4: Generate label set

{
  "required": ["type:feature", "priority:P1-High"],
  "recommended": ["agent:codegen", "phase:planning"],
  "optional": ["security", "cost-watch"],
  "automatic": ["state:pending", "hierarchy:leaf"]
}

📊 Output format

JSON format

{
  "issue_number": 270,
  "title": "XSS vulnerability in comment form",
  "analysis": {
    "type": "bug",
    "priority": "P0-Critical",
    "severity": "Sev.1-Critical",
    "estimated_time": "4-6 hours",
    "complexity": "medium",
    "agent_recommendation": "codegen + review",
    "escalation": "Guardian + CISO"
  },
  "labels": {
    "required": ["type:bug", "priority:P0-Critical"],
    "recommended": ["severity:Sev.1-Critical", "security"],
    "automatic": ["state:pending", "hierarchy:leaf"]
  },
  "rationale": {
    "type": "Keywords: 'vulnerability', 'XSS'",
    "priority": "Security issue with global impact",
    "severity": "Security vulnerability"
  }
}

Text format

🏷️ Issue Analysis Results

📋 Issue #270: XSS vulnerability in comment form

📊 Analysis:
- Type: bug (Keywords: 'vulnerability', 'XSS')
- Priority: P0-Critical (Security, global impact)
- Severity: Sev.1-Critical (Security vulnerability)
- Estimated: 4-6 hours
- Agent: codegen + review

🏷️ Labels:
✅ Required: type:bug, priority:P0-Critical
📎 Recommended: severity:Sev.1-Critical, security
🔄 Automatic: state:pending, hierarchy:leaf

⚠️ Escalation: Guardian + CISO (immediately)

🛡️ Error handling

Keyword detection failure

# デフォルトフォールバック
TYPE: type:feature (不明な場合)
PRIORITY: priority:P2-Medium (不明な場合)

Multiple TYPE matches

# 優先順位で決定
bug > security > feature > docs > test > chore

Conflicting labels

# 相互排他ラベルのチェック
state:pending ⊕ state:in-progress
priority:P0 ⊕ priority:P1

✅ Success criteria

Check item	Criteria
TYPE inference	Select 1
PRIORITY inference	Select 1
SEVERITY inference	Only for bugs
SPECIAL detection	Only when applicable
Provide rationale	Reason for each label
Escalation	Recommend when necessary

🔗 Related documents

Document	Purpose
`docs/LABEL_SYSTEM_GUIDE.md`	Label system details
`.github/labels.yml`	Label definitions
`agents/specs/coding/issue-agent.md`	IssueAgent specification
`docs/ENTITY_RELATION_MODEL.md`	Entity relationships

📝 Related Skills

Agent Execution: Agent execution after label assignment
Git Workflow: Commit after Issue resolution
Rust Development: Implementation quality check