Stock-Radar

Link: https://stock-radar-9t24-ppcrvodu4-darshans-projects-84a52e39.vercel.app/

Pay-per-use financial intelligence APIs with blockchain-native payments, async AI analysis jobs, and an operator-ready dashboard.

Why Stock-Radar

Stock-Radar is built for developers who want production-style market intelligence without subscription lock-in:

• Query only what you need, pay only for what you call.
• Expose discoverable machine-to-machine endpoints (/api/agent/discover, /api/agent/message).
• Support payment-gated routes through x402 + Aptos.
• Keep analysis state in Supabase and run compute in a dedicated Python backend.

What You Get

• Next.js app (web) with dashboard + public API routes.
• FastAPI backend (backend) for heavy analysis workloads.
• Async job flow for long-running analysis (/api/analyze + /api/analyze/status).
• Agent endpoints for momentum, scoring, sentiment, fundamentals, and orchestration.
• On-chain agent identity/reputation integration on Aptos testnet.

Architecture

flowchart LR
    U[User / Client] --> V[Vercel: Next.js Web]
    V -->|X-Backend-Api-Key| R[Railway: FastAPI Backend]
    V --> S[(Supabase)]
    R --> S
    R --> L[LLM Providers]
    R --> M[Market Data Providers]
    V --> A[Aptos Testnet / x402]

Loading

Repository Map

• web/: Next.js 16 app, UI, public API routes, x402 enforcement/proxy
• backend/: FastAPI service used by web for analysis/agent compute
• src/: Python analysis engine, agents, storage, orchestration logic
• move-agent-registry/: Move contracts for agent registry/reputation
• demo-client/: CLI demo for payment flow
• docs/deploy-vercel-railway.md: Production deployment runbook

Quick Start (Local)

1. Prerequisites

• Node.js 20+
• Python 3.11+
• npm
• Supabase project (URL + keys)

2. Configure Environment

From repo root:

cp .env.example .env

From web/:

cp .env.example .env.local

Minimum required values:

Backend (.env):

• BACKEND_API_KEY (optional for localhost-only dev; required for shared/prod environments)
• SUPABASE_URL
• SUPABASE_KEY
• one of ZAI_API_KEY or GROQ_API_KEY or GEMINI_API_KEY

Web (web/.env.local):

• NEXT_PUBLIC_SUPABASE_URL
• NEXT_PUBLIC_SUPABASE_ANON_KEY
• PY_BACKEND_URL=http://localhost:8000
• PY_BACKEND_API_KEY=<same as BACKEND_API_KEY> (optional when the backend is running on localhost)

3. Install Dependencies

# Python deps
pip install -r requirements.txt

# Web deps
cd web
npm install

4. Run Backend

From repo root:

uvicorn backend.app:app --host 0.0.0.0 --port 8000 --reload

Health check:

curl http://localhost:8000/health

5. Run Web App

In a second terminal:

cd web
npm run dev

Open:

• http://localhost:3000
• http://localhost:3000/x402-demo

API Quickstart

Service Health

curl -s http://localhost:3000/api/health | jq .

Agent Discovery

curl -s http://localhost:3000/api/agent/discover | jq .

Submit Async Analysis

curl -s -X POST http://localhost:3000/api/analyze \
  -H "Content-Type: application/json" \
  -d '{"symbol":"AAPL","mode":"intraday"}' | jq .

Poll Analysis Status

curl -s "http://localhost:3000/api/analyze/status?jobId=<JOB_ID>" | jq .

Ask Endpoint

curl -s -X POST http://localhost:3000/api/ask \
  -H "Content-Type: application/json" \
  -d '{"question":"Summarize AAPL momentum","symbol":"AAPL"}' | jq .

x402-Paid Endpoint Examples

Without payment header, paid routes return HTTP 402:

curl -i "http://localhost:3000/api/agent/momentum?symbol=AAPL"

For internal/testing bypass flows, configure INTERNAL_API_KEY and pass:

-H "X-Internal-Key: <INTERNAL_API_KEY>"

Deploy (Recommended: Vercel + Railway)

Use split deployment:

• web on Vercel
• backend on Railway (Dockerfile: Dockerfile.backend)