SkillAgentSearch skills...

Crypto Risk Assessment Tool

“A real-time, multi-chain crypto risk analysis system that integrates data retrieval and signal-based evaluation, using RAG to deliver strictly facts-only, non-advisory, and interpretable insights.”

Install / Use

npx skills add lraut-hub/Crypto-Risk-Assessment-Tool

Installs into whichever agent you are using.

About this skill

Quality Score

0/100

Supported Platforms

Universal

README

🕵️‍♂️ Crypto Risk Assessment Tool

FastAPI React Llama 3 ChromaDB

A high-precision, modular Retrieval-Augmented Generation (RAG) engine designed for deterministic crypto asset risk verification. This tool converts raw blockchain, market, and regulatory data into structured, source-backed signal reports.


🚀 The Problem

In the volatile crypto landscape, retail traders and compliance teams often struggle with:

  • Information Overload: Sifting through fragmented data across explorers and social media.
  • Speculative Bias: LLMs often "hallucinate" advice or provide subjective analysis.
  • Hidden Risks: Undetected honeypots, predatory taxes, or missing audits in obscure tokens.

The Solution: A fact-only, signal-first verification engine that prioritizes precision and transparency over conversational interaction.


🏗️ Architecture: The 9-Step Pipeline

The system follows a strict, deterministic workflow to ensure every risk report is grounded in verifiable data:

  1. Query Router: Classifies intent to block advisory/out-of-scope queries.
  2. Entity Resolver: Maps tickers/names to canonical IDs via CoinGecko.
  3. Asset Classifier: Identifies asset types (Smart Contract vs. Native Asset).
  4. Signal Engine: Concurrently fetches data from:
    • GoPlus Security: Smart contract vulnerabilities and tax flags.
    • CoinGecko: Market liquidity, rank, and exchange listings.
    • Regulatory Lists: Daily scraped data from FCA, SEC, and MAS.
  5. Signal Normalizer: Converts raw numbers into human-readable indicators (e.g., ⚠️ PREDATORY TAX).
  6. RAG Retriever: Pulls context from ChromaDB for signal grounding.
  7. LLM Explanation: Uses Llama 3.3 (Groq) to generate a 2-3 line factual summary.
  8. Output Formatter: Builds a structured Markdown/JSON report.
  9. Post-Guards: Final regex check to ensure no advisory language is present.

🛡️ Key Signal Layers

| Layer | Indicators Tracked | | :--- | :--- | | Structural | Honeypots, Minting permissions, Proxy/Upgradable contracts, Buy/Sell taxes. | | Market | Market Cap, Trading Volume, Exchange Listings, Market Rank. | | Audit | Security audit status, Verified source code, Known audit providers. | | Regulatory | Warnings from global regulators (FCA, SEC, MAS), Sanctioned status. |


🛠️ Tech Stack

  • Backend: FastAPI (Python 3.10+)
  • Frontend: React 19 + Vite + TypeScript + TailwindCSS
  • AI/LLM: Groq Cloud (llama-3.3-70b-versatile)
  • Vector DB: ChromaDB (Embeddings: BAAI/bge-large-en-v1.5)
  • Data APIs: GoPlus Security API, CoinGecko API

🏁 Getting Started

Prerequisites

  • Python 3.10+
  • Node.js 18+
  • API Keys: Groq, CoinGecko (optional), GoPlus

Installation

  1. Clone the Repository:
    git clone https://github.com/lraut-hub/Crypto-Risk-Assessment-Tool.git
    cd Crypto-Risk-Assessment-Tool
    
  2. Backend Setup:
    cd backend
    pip install -r requirements.txt
    python main.py
    
  3. Frontend Setup:
    cd ../frontend
    npm install
    npm run dev
    

⚠️ Disclaimer

Facts-only. No investment advice. This tool reports historical and technical signals; it does not predict future performance or guarantee safety. Always perform your own research (DYOR).


Built with precision for the next generation of crypto security.

Related Skills

View on GitHub
GitHub Stars7
CategoryDevelopment
Updated3mo ago
Forks0

Languages

Python

Security Score

67/100

Audited on Apr 28, 2026

No findings