An AI-native skill that scans your AI environment, quantifies your value across 4 dimensions, and generates a personalized career report — in 30 seconds.
Quick Start • How It Works • The Framework • Demo • 中文
Everyone is asking: "Will AI replace me?" — but nobody has a real answer.
In the AI era, professionals face an existential question: What is my real value when AI can do so much? Current assessments are either generic personality tests or hand-wavy "AI readiness" surveys. None of them actually look at how you use AI to evaluate where you stand.
DAMC is a Claude Code skill that uses AI to evaluate your relationship with AI. It automatically scans your actual AI usage environment — skills, configurations, memory systems, automation pipelines, git history — and combines this with a lightweight career profile to generate a quantified, actionable assessment.
No surveys. No self-reporting bias. Just data.
# Install via npx skills (recommended)
npx skills add https://github.com/anthropics/damc-skill
# Or clone manually
git clone https://github.com/anthropics/damc-skill.git ~/.claude/skills/damcIn Claude Code, simply type:
/damc
or say:
Evaluate my value in the AI era
That's it. DAMC handles the rest — scanning, scoring, and report generation — automatically.
DAMC v2 transforms from a one-time assessment into a continuous improvement system with three major new features:
Run /damc again and see how you've evolved. DAMC saves each scan locally to ~/.claude/damc-history/ and shows your delta changes on the next run.
📈 Progress Tracking (vs last scan 2026-05-10)
D Distillation 83 → 83 ↔️ No change
A Anti-Distill 76 → 75 ⬇️ -1
M AI Mastery 89 → 92 ⬆️ +3
C Career 81 → 83 ⬆️ +2
Overall: 82 → 83 ⬆️
Your HTML report now includes a Progress section with trend sparklines when you have 3+ scans.
DAMC doesn't just score you — it tells you exactly what to do. Based on your weakest sub-dimensions, it recommends specific skills to install, commands to run, and habits to build.
🎯 Smart Recommendations (based on your weakest sub-dimensions)
📉 Physical Presence (45/100) — A dimension gap
→ Join 1 offline tech community event per month
→ Try installing: meetup-finder, event-scheduler
📉 Memory System (13/25) — M dimension gap
→ You only have 3 memory types — add project and reference types
→ Run: claude memory add --type project "..."
📉 Standardization (72/100) — D dimension gap
→ You have 58 projects but few project-level CLAUDE.md files
→ Add dedicated CLAUDE.md for your top 5 projects
Every recommendation is concrete, data-backed, and immediately actionable — not generic advice.
找到你身边与 Agent 协作最6的人！ (Find the best Agent collaborator around you!)
Join a team with a group code and see how you rank against your colleagues or community members.
🏆 ACME-CORP Leaderboard (12 members)
1. 🥇 Jayden 83 pts AI Architect
2. 🥈 Alex 76 pts AI-Native Creator
3. 🥉 Sarah 71 pts Efficiency Machine
...
Your rank: #1 / 12 🎉
- Create or join a team with a simple code (e.g., "ACME-CORP", "UCWS-2026")
- Privacy-first: only your total score and archetype are shared with the team
- Optionally share your installed skills list with teammates
- Team leaderboard page:
vibergo.space/damc/team/{CODE}
The team feature turns DAMC into a viral loop — when one person shares their score, the whole team wants to try it.
┌───────────────────────────────────────────────────────────────────┐
│ DAMC v2 Workflow │
│ │
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ │
│ │ Phase 1 │─▶│ Phase 1.5 │─▶│ Phase 2 │─▶│ Phase 3 │ │
│ │ SCAN │ │ HISTORY │ │ PROFILE │ │ SCORE │ │
│ │ │ │ CHECK │ │ │ │ │ │
│ │ • Skills │ │ │ │ 3 Quick │ │ Scoring │ │
│ │ • Config │ │ • Load │ │ Questions │ │ Algorithm │ │
│ │ • Memory │ │ prev │ │ │ │ (22 sub- │ │
│ │ • Hooks │ │ scores │ │ • Role │ │ dimensions│ │
│ │ • MCP │ │ • Calc │ │ • Output │ │ mapped) │ │
│ │ • Git │ │ deltas │ │ • MBTI │ │ │ │
│ └──────────┘ └──────────┘ └──────────┘ └──────────┘ │
│ │ │
│ ▼ │
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ │
│ │ Phase 5 │─▶│ Phase 4.8 │◀─│ Phase 4 │◀─│Phase 3.5 │ │
│ │ UPLOAD │ │ TEAM │ │ REPORT │ │ SMART │ │
│ │ (optional)│ │LEADERBOARD│ │ │ │ RECS │ │
│ │ │ │ │ │ • HTML │ │ │ │
│ │ Scores + │ │ • Join │ │ • Radar │ │ • Weakest │ │
│ │ group_code│ │ team │ │ • Career │ │ dims │ │
│ │ → API │ │ • Show │ │ • Progress│ │ • Skills │ │
│ │ │ │ rank │ │ • Recs │ │ • Actions │ │
│ └──────────┘ └──────────┘ └──────────┘ └──────────┘ │
│ │
│ 💾 Save snapshot → ~/.claude/damc-history/scan-{date}.json │
└───────────────────────────────────────────────────────────────────┘
DAMC scans your Claude Code environment silently:
| Data Source | Signals Extracted |
|---|---|
~/.claude/CLAUDE.md |
Configuration depth, custom rules, workflow definitions |
~/.claude/skills/ |
Total count, self-built vs installed, category diversity |
~/.claude/memory/ |
Memory file count, type distribution |
settings.json |
Hooks count, MCP servers, permission model |
git log |
AI-collaborative commits (Co-Authored-By), commit frequency |
| System tools | Dev tool breadth (Node/Python/Go/Docker/K8s) |
- What's your role? (e.g., developer, PM, designer, marketer)
- What's your core output? (e.g., code, documents, designs, decisions)
- Your MBTI type? (optional — skip if you don't know)
Checks ~/.claude/damc-history/ for previous scan results. If found, calculates delta changes for each dimension and prepares progress tracking data.
22 sub-dimensions scored across 4 dimensions, producing an overall 0-100 score and one of 8 career archetypes.
Analyzes the 3 weakest sub-dimensions and generates specific, actionable recommendations — including exact skills to install, commands to run, and habits to build.
A beautiful, self-contained HTML report with:
- Radar chart visualization
- Animated score bars
- Career archetype analysis
- Progress tracking with trend sparklines (NEW)
- Personalized skill recommendations (NEW)
- Team leaderboard (NEW, if joined)
- Shareable format
Users can join a team by entering a group code. The platform returns a ranked leaderboard showing who has the highest DAMC score in the team.
Upload payload now includes optional group_code and display_name fields. API response includes leaderboard rank when a team is joined.
DAMC evaluates your value through four complementary lenses:
| Dimension | Full Name | Core Question | Weight |
|---|---|---|---|
| D | Distillation Value | Is your expertise worth distilling into an AI skill? | 25% |
| A | Anti-Distillation | Which of your abilities can AI never replicate? | 30% |
| M | AI Mastery | How well do you wield AI tools? | 25% |
| C | Career Compass | Based on D, A, M — where should you go next? | 20% |
Why A has the highest weight: In the AI era, your irreplaceable human qualities — creativity, emotional intelligence, cross-domain thinking, ambiguity tolerance — are your ultimate moat.
Based on High/Low combinations of D, A, M:
| Archetype | D | A | M | Description |
|---|---|---|---|---|
| 🏆 AI Architect | High | High | High | Top of the food chain — expert, irreplaceable, and AI-fluent |
| 🎯 Sleeping Expert | High | High | Low | Diamond-level expertise, just hasn't amplified it with AI yet |
| ⚡ Efficiency Machine | High | Low | High | High-output AI user, but the work itself is automatable |
| 🚨 Danger Zone | High | Low | Low | Replaceable knowledge + no AI skills = urgent action needed |
| 🌟 AI-Native Creator | Low | High | High | Value lies in creativity and judgment, AI amplifies it |
| 💎 Uncut Diamond | Low | High | Low | Irreplaceable soft skills, hasn't discovered AI leverage yet |
| 🔧 AI Tool Operator | Low | Low | High | Good at using AI, but "using tools" alone depreciates |
| 📚 Explorer | Low | Low | Low | Early career or transitioning — best time to reposition |
M Dimension (AI Mastery) — 100% automated, zero self-reporting:
| Sub-dimension | Max Score | Signals |
|---|---|---|
| Environment Config | 20 | CLAUDE.md depth, keybindings, settings |
| Skill Ecosystem | 25 | Skill count, self-built ratio, category coverage |
| Automation & Integration | 20 | Hooks, MCP servers |
| Memory System | 15 | Memory files, type diversity |
| Advanced Features | 20 | Multi-project config, cron, agent teams, AI commits |
D & A Dimensions — hybrid scoring: automated signals (40-70%) + role-based inference (30-60%)
C Dimension — derived from D, A, M with optional MBTI adjustment for career path recommendations
DAMC is designed with a strict privacy architecture:
✅ What stays local (ALWAYS):
• CLAUDE.md content
• Memory file content
• Git commit messages
• Skill names and paths
• Project paths
• Personal identifiers
❌ What NEVER leaves your machine:
• Raw scan data
• Configuration file contents
• Any personally identifiable information
🔒 What uploads (ONLY with explicit consent):
• Numeric scores only (D/A/M/C + 22 sub-scores)
• Archetype name
• Role (user-provided)
• Aggregate scan statistics (counts only, no names)
Users can choose:
- "Agree" → Full experience with platform features
- "Local mode" → Everything stays on your machine
- "Cancel" → Exit immediately
damc/
├── SKILL.md # Main skill definition (Claude Code format)
├── references/
│ ├── scoring-framework.md # 22-dimension scoring algorithm
│ └── career-archetypes.md # 8 archetype definitions + matching logic
├── templates/
│ └── report.html # Self-contained HTML report template
│ (CSS + JS + SVG radar chart, no dependencies)
├── docs/
│ ├── ARCHITECTURE.md # Technical deep-dive
│ └── assets/ # Demo screenshots and banner
├── README.md # This file (English)
├── README.zh-CN.md # Chinese version
└── LICENSE # MIT License
Local data (created at runtime):
~/.claude/damc-history/ # Progress tracking history
├── scan-2026-04-15.json # Timestamped score snapshots
├── scan-2026-05-10.json
└── scan-2026-05-25.json
-
Self-contained HTML report: Single file, no CDN dependencies, works offline. Dark theme with animated radar chart, gradient bars, and responsive design.
-
Hybrid scoring model: M dimension is 100% objective (automated scan), D and A blend automated signals with role-based inference, creating a balanced assessment that's neither pure black-box nor pure self-report.
-
Score-only upload: Privacy architecture ensures only numeric scores travel over the network. Raw content never leaves the machine.
-
Claude Code native: Built as a Claude Code skill — no separate backend, no browser extension, no additional setup. Just install and run.
DAMC is designed for global adoption:
- Language-agnostic scoring: The scan-based M dimension works regardless of language
- Role-universal framework: D, A, M, C dimensions apply to any profession in any country
- Cultural adaptability: Career archetypes map to universal workplace dynamics
- English + Chinese: Full bilingual support from day one
- Zero-infrastructure: Runs entirely within Claude Code — no servers needed for core functionality
| Tier | Features | Model |
|---|---|---|
| Free (LITE) | 4-dimension scores + archetype + local HTML report + progress tracking + smart recommendations | Open source skill |
| Pro | 22 sub-dimension deep dive + distillable skills list + moat analysis + 90-day action plan + team leaderboard | Platform (damc.ai) |
| Enterprise | Team analytics, department benchmarking, workforce transformation insights, org-wide DAMC rankings | B2B SaaS |
User runs /damc → Gets score + archetype → Shares with team
↓
Team members try /damc → Join team leaderboard → Compete & improve
↓
Re-scan shows progress → User shares improvement → More people join
↓
"找到你身边与 Agent 协作最6的人！" → Organic growth engine
# In Claude Code terminal
> /damc
🔒 DAMC Privacy Notice (please read)
I will scan your local environment to generate an Agent Health Report...
→ Type "agree" to continue
→ Type "local" for local-only mode
→ Type "cancel" to exit
> agree
🔍 Scanning environment...
✅ CLAUDE.md: 150 lines, 12 custom rules
✅ Skills: 83 total, 5 self-built, 10 categories
✅ Memory: 12 files, 4 types
✅ Hooks: 3 configured
✅ MCP: 8 servers
✅ Git: 45/200 AI-collaborative commits
Quick profile questions...
1. Your role: Frontend Developer
2. Core output: Code
3. MBTI: INTJ
📊 DAMC Assessment Complete
D ████████░░ 78 M █████████░ 85
A ██████░░░░ 62 C ██████░░░░ 65
Archetype: 🏆 AI Architect
"Deep expertise + irreplaceable skills + AI fluency = top of the food chain"
📈 Progress Tracking (vs last scan 2026-05-10)
D 78 → 78 ↔️ | A 60 → 62 ⬆️ +2 | M 82 → 85 ⬆️ +3 | C 63 → 65 ⬆️ +2
Overall: 71 → 72 ⬆️
🎯 Smart Recommendations
📉 Physical Presence (40/100) → Join 1 offline tech community event/month
📉 Memory System (12/25) → Run: claude memory add --type project "..."
📉 Cross-domain (68/100) → Try skills: seo-audit, geo, xiaohongshu
👥 Team: ACME-CORP — Rank #1 / 12 🎉
🔗 vibergo.space/damc/team/ACME-CORP
📄 Full report saved: ~/Desktop/DAMC-Report-2026-05-25.htmlContributions welcome! Areas we'd love help with:
- New language support: Translate archetypes and recommendations
- Scoring model refinement: Suggest new signals or weighting adjustments
- Platform integrations: VS Code, Cursor, Windsurf skill versions
- Enterprise features: Team-level analytics design
MIT License — see LICENSE for details.
Built for UCWS Singapore Hackathon 2026 — Skills Track
DAMC: Because in the AI era, knowing your value is the first step to multiplying it.