Workflow Analyzer

Workflow Analyzer Skill (Meta-Skill)

Purpose & Activation

What it does: Analyzes git history to detect workflow patterns and suggests skill improvements.

When it activates:

• "What patterns do you see in my commits?"
• "Analyze my workflow"
• "What skills should I create?"
• "What operations do I repeat?"

How it helps:

• Detects repeated manual operations
• Suggests new skills to create
• Identifies skill improvement opportunities
• Enables skill evolution based on actual usage

Key Operations

Operation 1: Commit Pattern Analysis

User Query:

"What patterns do you see?"

Confidence: 🟢 High (0.90)

• Exact keyword match: "patterns"
• Git repository available with commit history
• Clear intent to analyze workflow

Skill Response:

📊 Workflow Pattern Analysis (last 50 commits)

Repeated Operations Detected:
- "Update version.json" (8 times / 16%)
- "Fix typo in README" (6 times / 12%)
- "Update CHANGELOG" (7 times / 14%)
- "Sync version files" (8 times / 16%)

Pattern: Version management appears in 23/50 commits (46%)

Recommendation:
version-management skill already exists! ✅
This addresses your most frequent workflow.

Other opportunities:
- typo-detection skill (6 occurrences)
  Suggests: Auto-detect common typos before commit

Continue using version-management skill.
Consider typo-detection for future.

Why This Matters:

• 46% of commits are version-related → High-friction workflow that's already automated
• Typo fixes waste 12% of commit time → Opportunity for new skill
• Pattern detection reveals ROI: version-management saves ~800 tokens/week
• Quick win: Run ./scripts/claude-skills.sh show version-management to see usage patterns

Next Step: Want to analyze specific operation frequency or explore typo-detection skill creation?

Operation 2: Skill Improvement Suggestions

User Query:

"How can I improve my skills?"

Confidence: 🟢 High (0.85)

• Keyword match: "improve", "skills"
• Git history available for analysis
• Established skill ecosystem to analyze

Skill Response:

🔍 Skill Improvement Opportunities

Based on commit analysis:

version-management skill:
✅ Well-used (8 times in 50 commits)
💡 Could add: Auto-detect if sync-version.sh ran
💡 Could add: Suggest CHANGELOG template

commit-readiness-checker skill:
⚠️ Usage unknown (skill recently added)
📊 Monitor: Check if pre-commit queries increase

Recommendation:
Update version-management SKILL.md to include:
- sync-version.sh detection operation
- CHANGELOG template generation

This would save additional 200 tokens/release.

Why This Matters:

• Well-used skills (8+ times) deserve investment → High ROI for enhancements
• New skills need monitoring → Track adoption to validate value
• Incremental improvements compound → 200 tokens/release × 12 releases/year = 2,400 tokens saved
• Quick win: Edit .claude/skills/version-management/SKILL.md to add suggested operations

Next Step: Want detailed implementation guidance for these improvements?

Operation 3: New Skill Discovery

User Query:

"What skills should I create next?"

Confidence: 🟢 High (0.92)

• Direct question about skill creation
• Git history shows repeated patterns
• Clear data-driven recommendations possible

Skill Response:

🎯 Skill Recommendations (data-driven)

Based on your git history:

High Value:
1. typo-fixer skill
   Evidence: 6 "fix typo" commits in 2 weeks
   Savings: 300 tokens/query × 3 queries/week = 900 tokens/week

2. link-validator skill
   Evidence: 4 "fix broken link" commits
   Savings: 400 tokens/query × 2 queries/week = 800 tokens/week

Medium Value:
3. code-formatter skill
   Evidence: 3 formatting commits
   Savings: 200 tokens/query × 1 query/week = 200 tokens/week

Start with typo-fixer (highest ROI).

Why This Matters:

• Data-driven recommendations ensure high ROI → Build what you actually need
• Typo-fixer saves 900 tokens/week → 3,600 tokens/month (~$0.11/month)
• Pattern frequency indicates pain points → 6 typo commits in 2 weeks = clear need
• Quick win: Run ./scripts/skill-template-generator.sh typo-fixer to scaffold new skill

Next Step: Want help creating the typo-fixer skill SKILL.md structure?

Git-Based Analysis

Analyzes:

• Commit message patterns
• File change frequency
• Time between related commits
• Commit types (feat, fix, docs, etc.)

Detects:

• Repeated manual tasks
• High-friction workflows
• Missing automation opportunities

Meta-Skill: Enables Evolution

This skill answers: "Do skills improve automatically?"

Answer: No, but workflow-analyzer observes patterns and SUGGESTS improvements!

Example Evolution Loop:

Week 1: User commits "Update CHANGELOG" 3 times
Week 2: workflow-analyzer notices pattern
Week 3: Suggests: "Add CHANGELOG template to version-management skill"
Week 4: User updates skill with template
Week 5: CHANGELOG updates become faster

Token Efficiency

Without Skill: 900 tokens (manual git log analysis) With Skill: 400 tokens (structured git analysis) Savings: 500 tokens/query (56% reduction, projected)

Monthly: 2 analyses × 500 = 1,000 tokens (~$0.03/month)

See Also

• version-management skill
• commit-readiness-checker skill
• skill-creation-guide.md

Skill Version: 3.4.0 Last Updated: 2025-12-13 Type: Meta-skill (analyzes and improves other skills)