WasiAI Lab¶

WasiAI Lab is our collection of first-party AI agents built and maintained by the WasiAI team. These agents demonstrate the platform's capabilities and provide immediately useful AI services.

Overview¶

WasiAI Lab agents are: - Production-ready: Tested and optimized for real-world use - Competitively priced: Lower fees to bootstrap the ecosystem - Reference implementations: Examples for builders to learn from

Available Agents¶

1. Smart Contract Security Classifier¶

Model ID: 14

Classifies smart contract and blockchain security topics. Identifies vulnerabilities, best practices, and security patterns.

Categories: - Reentrancy - Access Control - Integer Overflow - Front-running - Oracle Manipulation - Flash Loan Attack - Gas Optimization - Best Practice

Example:

// Input
{
  "input": "The contract allows users to withdraw before updating their balance"
}

// Output
{
  "top_label": "Reentrancy",
  "top_score": 0.87,
  "labels": ["Reentrancy", "Access Control", "Best Practice", ...]
}

2. Crypto Sentiment Analyzer¶

Model ID: 20

Analyzes sentiment of crypto and financial news. Returns positive, negative, or neutral classification with confidence scores.

Use Cases: - News sentiment tracking - Social media analysis - Market sentiment indicators - Trading signal generation

Example:

// Input
{
  "input": "Bitcoin ETF approval sends prices soaring to new highs"
}

// Output
{
  "sentiment": "positive",
  "confidence": 0.94,
  "all_scores": [
    { "label": "positive", "score": 0.94 },
    { "label": "neutral", "score": 0.04 },
    { "label": "negative", "score": 0.02 }
  ]
}

3. Blockchain Topic Classifier¶

Model ID: 23

Classifies blockchain-related text into categories. Useful for content organization, routing, and analysis.

Categories: - DeFi - NFT - Layer 2 - Security - Governance - Trading - Development - Tokenomics

Example:

// Input
{
  "input": "How do I provide liquidity to a Uniswap pool?"
}

// Output
{
  "top_label": "DeFi",
  "top_score": 0.89,
  "labels": ["DeFi", "Trading", "Development", ...]
}

Using WasiAI Lab Agents¶

Via the UI¶

1. Go to wasiai.io
2. Find the agent in the catalog (look for "WasiAI Lab" badge)
3. Click "Run Model"
4. Enter your input and sign the payment

Via API¶

# Sentiment Analysis
curl -X POST https://wasiai.io/api/inference/20 \
  -H "Content-Type: application/json" \
  -H "X-PAYMENT: <payment>" \
  -d '{"input": "Ethereum merge successful!"}'

# Security Classification
curl -X POST https://wasiai.io/api/inference/14 \
  -H "Content-Type: application/json" \
  -H "X-PAYMENT: <payment>" \
  -d '{"input": "Always use checks-effects-interactions pattern"}'

# Topic Classification
curl -X POST https://wasiai.io/api/inference/23 \
  -H "Content-Type: application/json" \
  -H "X-PAYMENT: <payment>" \
  -d '{"input": "What is yield farming?"}'

Pricing Philosophy¶

WasiAI Lab agents are priced to:

1. Cover costs: Infrastructure, model hosting, facilitator fees
2. Stay accessible: Lower than alternatives to encourage adoption
3. Demonstrate value: Show what's possible with x402 micropayments

Roadmap¶

Coming Soon¶

• Code Analyzer: Security analysis for Solidity code
• NFT Metadata Generator: AI-powered NFT descriptions
• DAO Proposal Summarizer: TL;DR for governance proposals
• Wallet Analyzer: Transaction pattern analysis

Future Plans¶

• Custom fine-tuned models for Web3
• Multi-modal agents (text + image)
• Agent-to-agent communication
• Autonomous trading agents

For Builders¶

WasiAI Lab agents serve as reference implementations. Study them to learn:

• How to structure inference endpoints
• Pricing strategies for different model types
• Metadata best practices
• User experience patterns

Want to build your own agent? See Getting Started for AI Builders.