Skip to main content

Getting Started

Vishap is a 6-stage LLM pipeline that analyzes code, config, specs, logs, and data, and returns a structured Markdown report with risks, a STRIDE threat model, refactoring options, and an audit trail (SHA-256, PII detection, compliance flags, provenance).

Prerequisites​

  • Python 3.11 or 3.12
  • A DeepSeek / OpenAI-compatible API key, or a local Ollama server
  • Docker (optional, for the full stack)

Install​

python -m venv .venv
# Windows
.\.venv\Scripts\Activate.ps1
# Linux / macOS
source .venv/bin/activate

pip install -r requirements.txt

Copy .env.example to .env and fill in your keys.

Quick start​

CLI​

vishap analyze path/to/file.py

API​

python -m src.main
# -> http://localhost:5001/docs
curl -X POST http://localhost:5001/analyze \
  -H "Content-Type: application/json" \
  -d '{"material": "def f(): pass", "filename": "x.py"}'

Async (Celery + Redis)​

docker compose up -d redis worker api

curl -X POST http://localhost:5001/analyze/async \
  -H "Content-Type: application/json" \
  -d '{"material": "def f(): pass", "filename": "x.py"}'
# -> {"job_id": "...", "status": "queued"}

curl http://localhost:5001/status/<job_id>
curl http://localhost:5001/report/<job_id>

Docker​

docker compose up -d
# API  -> http://localhost:5001
# UI   -> http://localhost:8501

Next steps​