hackathon-repo-bootstrap（黑客马拉松回购引导程序）

hackathon-repo-bootstrap

Goal

Generate a ready-to-run project scaffold for a hackathon project, configured for the recommended stack (Next.js + Render + Supabase + LLM router), with environment variables, deployment configuration, and folder structure ready for immediate development.

Trigger Conditions

Use this skill when:

• MVP scope is locked and the tech stack is confirmed from hackathon-task-planner
• The team is about to begin implementation and needs a project skeleton
• A deployment target (Vercel, Render, Railway) has been chosen
• Environment variables and API keys need to be structured before coding begins
• Invoked once at the start of Phase 5 (Build), immediately before the first hackathon-code-implementer call

Inputs

Outputs

Rules

1. Generate a monorepo structure with frontend/ and backend/ subdirectories when team_size >= 3; otherwise generate a single Next.js full-stack app.
2. Include a README.md stub with setup instructions in the scaffold root.
3. All API keys must appear as environment variables — never hardcoded.
4. env_template must list every required variable with a descriptive comment.
5. startup_commands must be executable from a clean machine with Node and Python installed.
6. deployment_config must include both a local run config and a platform deploy config.
7. Flag any required external service (Supabase, Groq, etc.) with a [REQUIRES ACCOUNT] note.

Output Format

directory_structure: |
  <project-name>/
  ├── frontend/
  │   ├── ...
  ├── backend/
  │   ├── ...
  └── README.md

env_template: |
  # <project-name> Environment Variables
  # Frontend
  NEXT_PUBLIC_API_URL=http://localhost:8000
  # Backend
  <VAR_NAME>=<description>   # [REQUIRES ACCOUNT]

startup_commands:
  - step: <number>
    command: "<shell command>"
    description: "<what it does>"

deployment_config:
  - platform: "<Vercel|Render|Railway>"
    file: "<config file name>"
    content: |
      <config content>

llm_routing_scaffold:
  provider: "<groq|openrouter|nvidia-nim|openai>"
  language: "<python|typescript>"
  snippet: |
    <code>

database_scaffold:
  provider: "<supabase|postgres|sqlite>"
  language: "<python|typescript>"
  snippet: |
    <code>

recommended_skills:
  - "<skill-name>"

Example

Input:

project_title: "AnchorAI"
tech_stack: ["Next.js", "FastAPI", "Supabase", "Groq"]
mvp_features:
  - feature: "GPT-4 emotional check-in conversation"
  - feature: "Session memory"
  - feature: "Crisis escalation (mocked)"
deployment_targets:
  - role: "frontend"
    platform: "Vercel"
  - role: "backend"
    platform: "Render"
llm_provider: "groq"
database: "supabase"
team_size: 3

Output:

directory_structure: |
  anchorai/
  ├── frontend/                  # Next.js 14 app
  │   ├── app/
  │   │   ├── page.tsx
  │   │   └── api/
  │   ├── components/
  │   ├── .env.local.example
  │   ├── next.config.js
  │   └── package.json
  ├── backend/                   # FastAPI service
  │   ├── main.py
  │   ├── routers/
  │   │   └── chat.py
  │   ├── services/
  │   │   ├── llm.py
  │   │   └── memory.py
  │   ├── .env.example
  │   └── requirements.txt
  └── README.md

env_template: |
  # AnchorAI — Backend Environment Variables
  GROQ_API_KEY=                  # [REQUIRES ACCOUNT] https://console.groq.com
  SUPABASE_URL=                  # [REQUIRES ACCOUNT] https://supabase.com
  SUPABASE_ANON_KEY=             # [REQUIRES ACCOUNT]
  DEMO_USER_ID=demo

  # AnchorAI — Frontend Environment Variables
  NEXT_PUBLIC_API_URL=http://localhost:8000

startup_commands:
  - step: 1
    command: "cd anchorai/backend && python -m venv .venv && source .venv/bin/activate && pip install -r requirements.txt"
    description: "Install Python dependencies"
  - step: 2
    command: "cp .env.example .env && # fill in API keys"
    description: "Configure environment variables"
  - step: 3
    command: "uvicorn main:app --reload --port 8000"
    description: "Start FastAPI backend"
  - step: 4
    command: "cd anchorai/frontend && npm install && cp .env.local.example .env.local && npm run dev"
    description: "Start Next.js frontend"

deployment_config:
  - platform: "Render"
    file: "render.yaml"
    content: |
      services:
        - type: web
          name: anchorai-api
          runtime: python
          buildCommand: "pip install -r requirements.txt"
          startCommand: "uvicorn main:app --host 0.0.0.0 --port $PORT"
          envVars:
            - key: GROQ_API_KEY
              sync: false
            - key: SUPABASE_URL
              sync: false
  - platform: "Vercel"
    file: "vercel.json"
    content: |
      {
        "env": {
          "NEXT_PUBLIC_API_URL": "@anchorai-api-url"
        }
      }

llm_routing_scaffold:
  provider: "groq"
  language: "python"
  snippet: |
    from groq import Groq

    client = Groq(api_key=os.environ["GROQ_API_KEY"])

    def chat(messages: list[dict], model: str = "llama3-70b-8192") -> str:
        response = client.chat.completions.create(
            model=model,
            messages=messages,
        )
        return response.choices[0].message.content

database_scaffold:
  provider: "supabase"
  language: "python"
  snippet: |
    from supabase import create_client

    supabase = create_client(
        os.environ["SUPABASE_URL"],
        os.environ["SUPABASE_ANON_KEY"]
    )

    def save_session_summary(user_id: str, summary: str):
        supabase.table("sessions").upsert({
            "user_id": user_id,
            "summary": summary
        }).execute()

    def get_session_summary(user_id: str) -> str:
        result = supabase.table("sessions").select("summary").eq("user_id", user_id).execute()
        return result.data[0]["summary"] if result.data else ""

recommended_skills:
  - "hackathon-code-implementer"
  - "hackathon-risk-analyzer"

Context Files

Knowledge Base

• knowledge/hackathon-reference-architecture.md
• knowledge/hackathon-tools.md
• knowledge/hackathon-mvp-strategy.md
• knowledge/hackathon-common-failures.md

Playbooks

• playbooks/hackathon-workflow.md
• playbooks/24h-hackathon-playbook.md
• playbooks/36h-hackathon-playbook.md
• playbooks/48h-hackathon-playbook.md