# Veil Armor - Requirements

# Complete implementation of all 10 OWASP LLM Vulnerabilities

# ===== Core Framework =====

python-dotenv>=1.0.0

pydantic>=2.0.0

toml>=0.10.2

pyyaml>=6.0.0

# ===== Web Framework =====

fastapi>=0.104.0

uvicorn[standard]>=0.24.0

python-multipart>=0.0.6

slowapi>=0.1.9 # LLM10: Rate limiting

starlette>=0.27.0

# ===== LLM API (Gemini) =====

google-genai>=1.0.0 # New Gemini SDK (replaces google-generativeai)

# ===== LLM01: Prompt Injection Detection =====

# Note: Vigil must be installed from GitHub, not PyPI

# Install separately with: pip install git+https://github.com/deadbits/vigil-llm.git

# ===== LLM02: Sensitive Information Disclosure =====

presidio-analyzer>=2.2.0 # Microsoft PII detection with NER

presidio-anonymizer>=2.2.0 # PII redaction

spacy>=3.7.0 # NLP engine for Presidio

# After install, download model: python -m spacy download en_core_web_lg

# ===== LLM03: Supply Chain Vulnerabilities =====

# Note: Trivy CLI must be installed separately (system package)

# Windows: choco install trivy OR download from https://github.com/aquasecurity/trivy/releases

# Linux: apt-get install trivy OR brew install trivy

# ===== LLM04: Data/Model Poisoning =====

scikit-learn>=1.3.0 # Anomaly detection (IsolationForest)

# ===== LLM05: Improper Output Handling =====

guardrails-ai>=0.6.0 # Output validation and sanitization

bleach>=6.0.0 # HTML/XSS sanitization

# ===== LLM06: Excessive Agency =====

# Implemented in core logic (action limiting, no external deps)

# ===== LLM07: System Prompt Leakage =====

# Canary tokens (built-in to core, uses SQLite)

# ===== LLM08: Vector/Embedding Weaknesses =====

chromadb>=0.4.0 # Vector database for RAG

sentence-transformers>=2.2.0 # Local embeddings

# ===== LLM09: Misinformation =====

# Fact-checking with Gemini API confidence scoring

# ===== ML and NLP =====

transformers>=4.30.0

torch>=2.0.0

numpy>=1.26.0 # Fixed: 1.26+ required for Python 3.12+ (distutils removed)

# ===== YARA Pattern Matching =====

yara-python>=4.3.0

# Note: System YARA must be installed BEFORE yara-python

# Windows: Download from https://github.com/VirusTotal/yara/releases/tag/v4.3.2

# Linux: apt-get install yara

# macOS: brew install yara

# ===== Utilities =====

requests>=2.31.0

aiohttp>=3.9.0

pandas>=2.0.0

python-jose[cryptography]>=3.3.0 # JWT tokens

nltk>=3.8.0 # Natural language toolkit (Vigil dependency)

# ===== Logging and Monitoring =====

loguru>=0.7.0

prometheus-client>=0.19.0

# ===== Testing =====

pytest>=7.4.0

pytest-asyncio>=0.21.0

pytest-cov>=4.1.0

httpx>=0.25.0

# ===== Redis (optional for distributed rate limiting) =====

redis>=5.0.0

# ===== Database =====

sqlalchemy>=2.0.0 # For canary token storage

# ===== Development =====

black>=23.0.0

flake8>=6.1.0

mypy>=1.5.0

pre-commit>=3.4.0