AI Identification Card — Test your AI Native level.
Real AI proficiency cards — generated from public X data
Hermes ID is an AI capability intelligent assessment tool powered by Hermes Agent.
Enter any X/Twitter handle. Hermes Agent automatically scrapes and deeply analyzes the account's recent public tweets, then outputs a total AI Score (0–100), a multi-dimensional proficiency profile, and a shareable card image — all in one step.
One-line pitch: For the AI era — test how AI Native you really are.
- Satisfy the urge to show off, self-reflect, and compare
- Rapidly build Hermes Agent's presence in the AI community
- A viral-ready micro-product, easy for KOLs to share and attract new users
@teknium → Scrape tweets → LLM analysis → Generate card → Share on X
- Input — Provide any X handle (your own or anyone else's)
- Scrape — Hermes Agent collects 30–90 days of recent public tweets
- Analyze — LLM scores the user across 8 AI proficiency dimensions with evidence-based reasoning
- Generate — Python generates a card image with halftone avatar, radar chart, progress bars, and barcode
- Share — Get the card + one-click share link for X with Twitter Card preview
Every score is evidence-based — backed by specific tweet content. No speculation, no inflation.
| Dimension | Weight | What It Measures |
|---|---|---|
| AI Usage | 15% | Frequency and breadth of actual AI tool use |
| AI Understanding | 15% | Depth of technical AI/ML knowledge |
| Communication | 12% | Quality of AI community engagement |
| Product Building | 15% | Actually shipping things built with AI |
| Adoption Speed | 12% | How fast they try new AI releases |
| Prompt Engineering | 10% | Sophistication of prompting techniques |
| Critical Awareness | 10% | Understanding of AI limitations and risks |
| Knowledge Sharing | 11% | Teaching and educating others about AI |
Total Score = weighted average across all dimensions.
| Score | Level | Description |
|---|---|---|
| 95–100 | Agent God | Pinnacle of AI mastery. Builds, teaches, innovates, and shapes the field. |
| 88–94 | AI Native | AI is fully integrated into their identity and work. |
| 78–87 | AI Strategist | Strong, well-rounded AI skills. Clearly ahead of the curve. |
| 65–77 | AI Explorer | Active AI user and learner. Solid foundation, room to grow. |
| 50–64 | AI Curious | Beginning their AI journey. Interested but limited depth. |
| 35–49 | AI Aware | Some awareness of AI, limited engagement. |
| 0–34 | AI Normie | Little to no AI engagement visible in public content. |
Each generated card includes:
- Real X profile picture with manga-style halftone filter
- 8-axis radar chart for visual dimension overview
- Progress bars for detailed score breakdown
- AI Score, Level, and Role assignment
- Personalized summary — a shareable one-liner
- Unique barcode per handle
- Share CTA — "Tag @Hermes_ID and challenge your friends"
git clone https://github.com/hermes-intel/hermes-agent.git
cd hermes-agent/skills/social-media/hermes-id
pip install Pillow
# Generate a test card with sample scores
python3 test_e2e.py teknium --card-only
# → Card saved: /tmp/hermes_id_teknium.png# Set one of these API keys:
export OPENAI_API_KEY="sk-..."
# or: export OPENROUTER_API_KEY="sk-..."
# or: export ANTHROPIC_API_KEY="sk-..."
# Run the complete pipeline: scrape → analyze → generate
python3 test_e2e.py teknium
# Try any X handle
python3 test_e2e.py elonmusk
python3 test_e2e.py karpathy# Create a score JSON
cat > /tmp/score.json << 'EOF'
{
"handle": "yourhandle",
"total_score": 85,
"level": "AI Strategist",
"role": "AI Architect",
"summary": "A hands-on builder shipping AI tools at the frontier.",
"dimensions": [
{"name": "AI Usage", "score": 90},
{"name": "AI Understanding", "score": 88},
{"name": "Communication", "score": 75},
{"name": "Product Building", "score": 92},
{"name": "Adoption Speed", "score": 85},
{"name": "Prompt Engineering", "score": 78},
{"name": "Critical Awareness", "score": 70},
{"name": "Knowledge Sharing", "score": 82}
]
}
EOF
# Generate the card
python3 skills/social-media/hermes-id/references/generate_card.py /tmp/score.json /tmp/my_card.pngInstall the skill in Hermes Agent and simply chat:
> @teknium
> Score @elonmusk's AI level
> Generate an AI ID card for @karpathy
Hermes Agent will automatically activate the Hermes ID skill, scrape tweets, run LLM analysis, generate the card, and present results with a share link.
hermesid.wtf — The public-facing website with:
- Interactive card gallery featuring AI leaders
- "Get Your Card" — enter any X handle to generate a card instantly
- Share on X with rich Twitter Card previews (card image appears in the tweet)
- Download card as PNG for manual sharing
- Cyberpunk aesthetic with ASCII rain, 3D card flip animations, glowing aura
cd site/
pip install -r requirements.txt
# Run the card API server
python card_api.py
# → http://127.0.0.1:5050
# Test
curl http://127.0.0.1:5050/api/card?handle=teknium -o test.pngSee site/README.md for full deployment instructions (nginx, SSL, systemd).
| Endpoint | Description |
|---|---|
GET /api/card?handle=xxx |
Generate and return a PNG card image |
GET /card/<handle> |
HTML page with Twitter Card meta tags for rich link previews |
GET /api/health |
Health check |
hermes-agent/
├── skills/social-media/hermes-id/ # Core Hermes ID Skill
│ ├── SKILL.md # Skill definition (workflow + LLM prompt)
│ ├── README.md # Skill documentation
│ ├── test_e2e.py # End-to-end pipeline test
│ ├── requirements.txt # Python dependencies
│ └── references/
│ ├── generate_card.py # Card image generator (PIL/Pillow)
│ └── scoring-rubric.md # Detailed scoring criteria
├── site/ # Website (hermesid.wtf)
│ ├── index.html # Landing page
│ ├── card_api.py # Flask API for dynamic card generation
│ ├── favicon.png # Site icon
│ ├── requirements.txt # API dependencies
│ ├── README.md # Deployment guide
│ └── cards/ # Pre-generated example cards
│ ├── teknium.png
│ ├── karpathy.png
│ ├── sama.png
│ ├── elonmusk.png
│ └── emaboringkass.png
└── ... # Hermes Agent core (upstream)
| Component | Technology | Purpose |
|---|---|---|
| Skill Engine | Hermes Agent + SKILL.md | Workflow orchestration, browser scraping, tool coordination |
| LLM Analysis | OpenAI / OpenRouter / Anthropic | Multi-dimensional scoring with evidence-based reasoning |
| Card Generator | Python + Pillow (PIL) | Pixel-perfect card image rendering |
| Avatar Service | unavatar.io | Free X profile picture fetching (no API key needed) |
| Website | HTML/CSS/JS + Canvas API + Web Audio API | Landing page with interactive animations |
| Card API | Flask + Gunicorn | Dynamic card generation + Twitter Card meta tags |
| Hosting | Nginx + Let's Encrypt | Static serving + API proxy + SSL |
- Cards include CTA: "Tag @Hermes_ID and challenge your friends — test any @handle!"
- Share links come pre-filled with tweet text mentioning @Hermes_ID
- Twitter Card meta tags ensure card images appear as rich previews in tweets
- Seed strategy: Score AI KOLs and @ them to trigger first wave of shares
- Every card is unique and personalized — high share motivation
Primary: AI enthusiasts, developers, KOLs, researchers, agent builders, tech bloggers, product managers, founders
Secondary: Recruiters (quick AI capability assessment), community managers (identify high-potential builders), investors (evaluate candidates)
git clone https://github.com/hermes-intel/hermes-agent.git
cd hermes-agent
pip install Pillow # for card generationSee the Contributing Guide for development setup and PR process.
- hermesid.wtf — Try it now
- @Hermes_ID on X — Follow for updates
- GitHub Issues — Bug reports and feature requests
MIT — see LICENSE.
Powered by Hermes Agent.