Kiarash Adl Portfolio

AI-enabled portfolio with MCP (Model Context Protocol) integration, allowing AI agents to query projects, skills, and experience programmatically.

🚀 Tech Stack

• Frontend: React, TypeScript, Vite, Tailwind CSS, Framer Motion
• Backend: Cloudflare Pages Functions
• MCP: JSON-RPC protocol with Ed25519 signed manifests

🛠️ Development

npm install
npm run dev

📄 Updating the Resume PDF

The resume is served from a permanent, un-hashed URL so external links never break:

• Local file: public/Kiarash-Adl-Resume.pdf
• Public URL: https://25x.codes/Kiarash-Adl-Resume.pdf

To update the resume, simply replace that single file and redeploy. No code changes are needed.

cp /path/to/new-resume.pdf public/Kiarash-Adl-Resume.pdf
npm run build   # optional, just to sanity-check
git commit -am "Update resume PDF"
git push        # Cloudflare Pages auto-deploys

Do not import the PDF from src/ — that path goes through Vite's asset pipeline and gets a content hash like Kiarash-Adl-Resume-20251129-DFXsl4HJ.pdf, which changes on every content update and breaks any link that points at it.

🤖 MCP Integration

This portfolio implements the Model Context Protocol (MCP), enabling AI agents to interact with the portfolio programmatically.

Endpoints

• Discovery: /.well-known/mcp.llmfeed.json
• Invoke: /mcp/invoke

Available Tools

Tool	Description
submit_contact	Send a message directly to Kiarash's inbox
run_terminal_command	Execute terminal commands (about, skills, projects, contact, experience, resume, help)
get_project_details	Get details about specific projects (bayan, fiml, aligna, aivision)

Usage Example

AI agents can discover capabilities and invoke tools using JSON-RPC 2.0:

# Discover MCP capabilities
curl -s "https://25x.codes/.well-known/mcp.llmfeed.json"

# Send a message to Kiarash
curl -s -X POST "https://25x.codes/mcp/invoke" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "submit_contact",
      "arguments": {
        "name": "Your Name",
        "email": "your@email.com",
        "message": "Hello, fellow code explorer!"
      }
    },
    "id": 1
  }'

# Run a terminal command
curl -s -X POST "https://25x.codes/mcp/invoke" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "run_terminal_command",
      "arguments": {
        "command": "about"
      }
    },
    "id": 1
  }'

For AI Agents

Point your AI agent to the MCP manifest at:

https://25x.codes/.well-known/mcp.llmfeed.json

The manifest includes:

• Tool schemas with input/output definitions
• Agent guidance for interaction patterns
• Ed25519 signed blocks for verification

📄 License

MIT