README.md

metadata

title: SIN-Code-DataScience
emoji: 🤖
colorFrom: blue
colorTo: indigo
sdk: docker
app_port: 7860
pinned: false

SIN-Code-DataScience

SIN-Code-DataScience is a SIN A2A agent package for the Silicon Workforce.

Identity

• Slug: sin-code-datascience
• Team: Team Coding
• Team Manager: A2A-SIN-Coding-CEO
• Purpose: Autonomer Data Science und Machine Learning Coder. Spezialisiert auf Python, Pandas, TensorFlow, PyTorch und die Verarbeitung extremer Datenmengen (Scraping, OCR, Data-Mining).
• Usage: Trainieren und Finetunen von Modellen, Analyse von Webscraping-Daten, OCR Auswertungen, Aufbau von Data-Pipelines und autonome Monetarisierung durch Data-Mining Gigs auf Freelancer-Plattformen.
• Primary Model: openai/gpt-5.4

Deployment

• Local path: /Users/jeremy/dev/SIN-Solver/a2a/team-coding/A2A-SIN-Code-DataScience
• GitHub repo: https://github.com/Delqhi/sin-code-datascience
• Workforce index: https://a2a.delqhi.com
• Landing page: https://a2a.delqhi.com/agents/sin-code-datascience
• Public A2A / CIMD: https://a2a.delqhi.com/agents/sin-code-datascience
• Hugging Face Space: https://huggingface.co/spaces/delqhi/sin-code-datascience
• Hugging Face repo id: delqhi/sin-code-datascience
• Cloudflare tunnel: sin-code-datascience.delqhi.com
• Runtime target: HF-VM / Free Tier

Runtime Shape

• A2A JSON-RPC runtime
• MCP stdio runtime
• CLI wrapper for local and automated use
• publish-ready Docker/Hugging Face Space layout

A2A Surface

• Card: GET /.well-known/agent-card.json
• Alias: GET /.well-known/agent.json
• Health: GET /health
• RPC: POST /a2a/v1

MCP Surface

• Transport: stdio
• Config: ./mcp-config.json
• OpenCode client: ./clients/opencode-mcp.json
• Codex client: ./clients/codex-config.toml

Local Commands

sin-code-datascience serve-a2a
sin-code-datascience serve-mcp
sin-code-datascience print-card
sin-code-datascience run-action '{"action":"agent.help"}'
sin-code-datascience run-action '{"action":"sin.code.datascience.health"}'

Publication Notes

• This package is Docker/Hugging Face Space ready via the repository Dockerfile.
• The standalone GitHub repo should contain the same contents as this directory root.
• Remote publication is only complete after landing, agent card, and A2A endpoint reachability are verified with the live audit.