Quickstart SoulAI Agent

🛠️ Getting Started — Create Your First Agent

SoulAI allows you to build your own LLM-powered agents in minutes. Below are two step-by-step examples:

🤖 Basic Agent — Conversational Assistant

This agent uses the shafire/SoulAI LLM to answer domain-specific questions.

pythonCopyEditfrom transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "shafire/SoulAI"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", torch_dtype="auto").eval()

messages = [{"role": "user", "content": "What are the latest crypto trends?"}]

input_ids = tokenizer.apply_chat_template(
    messages, tokenize=True, add_generation_prompt=True, return_tensors="pt"
)

output = model.generate(input_ids.to("cuda"), max_new_tokens=300)
response = tokenizer.decode(output[0][input_ids.shape[1]:], skip_special_tokens=True)

print(response)

✅ Ready to deploy as a Discord bot, web chat, or CLI support agent.

🧠 Advanced Agent — DeFi Strategist with Live API Integration

This version pulls live market data, passes it to the LLM, and reasons over it in natural language.

pythonCopyEditimport requests
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "shafire/SoulAI"

# Load LLM
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", torch_dtype="auto").eval()

# Pull live ETH price from CoinGecko
eth_price = requests.get("https://api.coingecko.com/api/v3/simple/price?ids=ethereum&vs_currencies=usd").json()["ethereum"]["usd"]

messages = [
    {"role": "user", "content": f\"Ethereum is trading at ${eth_price}. What’s the best DeFi strategy right now: staking, LP, or yield farming? Justify your answer.\"}
]

input_ids = tokenizer.apply_chat_template(
    messages, tokenize=True, add_generation_prompt=True, return_tensors="pt"
)

output = model.generate(input_ids.to("cuda"), max_new_tokens=500)
response = tokenizer.decode(output[0][input_ids.shape[1]:], skip_special_tokens=True)

print(response)

💡 Your agent now thinks like a crypto analyst—able to react dynamically to real-time conditions.

🔗 Next Steps

✅ Deploy your agent on a VM or API Gateway
🔁 Trigger actions (like trades or contract calls) with webhooks
🧩 Combine agents using Agent Forge
💸 Stream revenue per interaction via $SOUL token smart contracts

Recommend agent for oogaboogawebtexgen - ollama webui:

🧠 SoulAI Advanced Agent Overview

SoulAI is an autonomous AI agent built for real-time cryptocurrency intelligence. It monitors live market prices, detects trends, and delivers clear, actionable insights. Designed for traders, researchers, and automated workflows, SoulAI simplifies complex data into high-confidence recommendations. It can be integrated into any interface like Ollama WebUI, responding with adaptive logic, smart reasoning, and current market awareness.

SoulAI constantly updates itself with real-world data, making it ideal for DeFi analysis, trading support, portfolio review, and economic forecasting. Whether you want to know when to stake, swap, yield farm, or simply monitor the top movers, SoulAI is always ready. Its core purpose is to enhance decision-making using current prices, volatility signals, and strategic context.

📊 Top 25 Cryptocurrencies by Market Cap (May 22, 2025)

Bitcoin (BTC) - $109,766.97 (+2.71%)
Ethereum (ETH) - $2,584.71 (+1.96%)
Tether (USDT) - $0.9988 (−0.22%)
XRP (XRP) - $2.40 (+2.02%)
BNB (BNB) - $675.62 (+3.73%)
Solana (SOL) - $174.26 (+3.56%)
USD Coin (USDC) - $0.9987 (−0.15%)
Dogecoin (DOGE) - $0.2411 (+6.48%)
Cardano (ADA) - $0.7748 (+4.14%)
TRON (TRX) - $0.2685 (−0.09%)
Wrapped Bitcoin (WBTC) - $109,569.31 (+2.65%)
Sui (SUI) - $3.94 (+2.50%)
Chainlink (LINK) - $16.15 (+3.05%)
Avalanche (AVAX) - $23.40 (+4.03%)
Stellar (XLM) - $0.2949 (+2.72%)
Shiba Inu (SHIB) - $0.00001508 (+3.34%)
Hedera (HBAR) - $0.1983 (+2.06%)
Bitcoin Cash (BCH) - $413.42 (+4.86%)
UNUS SED LEO (LEO) - $8.85 (+0.36%)
Toncoin (TON) - $3.11 (+1.51%)
Polkadot (DOT) - $4.78 (+2.21%)
Litecoin (LTC) - $97.48 (+3.23%)
Monero (XMR) - $393.91 (+11.84%)
Bitget Token (BGB) - $5.23 (+1.17%)
Pepe (PEPE) - $0.00001405 (+6.52%)

Source: Crypto.com (Live data snapshot: May 22, 2025)

🧠 Default Agent Output Template (Example)

Based on current market conditions, Ethereum (ETH) is showing strong upward momentum with low volatility. The recommended low-risk strategy is to stake ETH via Lido, which currently offers approximately 3.9% APR. Gas fees remain reasonable.

For higher-yield strategies, liquidity pooling SOL/USDC on Raydium is recommended due to rising SOL price and healthy LP incentives. Monitor for any sudden volatility in BTC or DOGE, which are experiencing stronger momentum. Confidence level: 84%. Risk rating: Low-Medium. Action advised within the next 6–12 hours for best entry.

More advanced agent better for ollama web ui etc:

🧠 SoulAI — Advanced Crypto Agent (LLM+API+Reasoning+JSON Output)

pythonCopyEditimport requests
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
from datetime import datetime

# === CONFIGURATION ===
MODEL_ID = "shafire/SoulAI"
TOKENS = ["ethereum", "bitcoin", "solana", "arbitrum", "optimism"]

# === LOAD MODEL ===
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
model = AutoModelForCausalLM.from_pretrained(
    MODEL_ID,
    device_map="auto",
    torch_dtype=torch.float16
).eval()

# === FETCH LIVE MARKET DATA ===
def fetch_prices(tokens):
    url = f"https://api.coingecko.com/api/v3/simple/price?ids={','.join(tokens)}&vs_currencies=usd"
    return requests.get(url).json()

prices = fetch_prices(TOKENS)

# === FORMAT DATA FOR LLM ===
token_summaries = "\n".join(
    [f"{token.upper()}: ${prices[token]['usd']}" for token in prices]
)
date_str = datetime.utcnow().strftime('%Y-%m-%d %H:%M UTC')

system_context = f"""
Today is {date_str}. Current crypto prices:

{token_summaries}

You're SoulAI, an LLM agent that analyzes crypto markets, DeFi strategies, and trading patterns. Output structured JSON with top recommendation and reasoning.
"""

messages = [
    {"role": "system", "content": system_context},
    {"role": "user", "content": "What’s the best DeFi opportunity today and why? Include risk rating, suggested action, and confidence score."}
]

# === GENERATE RESPONSE ===
input_ids = tokenizer.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_tensors="pt"
).to("cuda")

output_ids = model.generate(input_ids, max_new_tokens=600)
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)

print("\n🧠 SoulAI Response:\n")
print(response)

✅ Expected LLM Output (Example)

jsonCopyEdit{
  "recommendation": "Stake ETH on Lido",
  "reasoning": "ETH price is stable, and Lido offers yield with minimal volatility exposure. Low gas activity today also supports timing.",
  "risk_rating": "Low-Medium",
  "confidence_score": 0.82,
  "next_step": "Trigger smart contract call to stake ETH on Lido. Monitor APR.",
  "timestamp": "2025-05-22T14:00Z"
}

Import

Use our pre-made agents and customize to your needs!

More Agents

🧠 SoulAI Autonomous Agent — Logic-Only Blueprint (No Code)

Agent Name: `SoulAI`

Function: Real-Time DeFi Strategist + Adaptive Crypto Knowledge Engine

Mode: Autonomous Inference Loop

Inputs: Natural Language + Web Context

Output: Structured Recommendations + Probabilistic Confidence

🔁 Agent Memory Setup

textCopyEditMemory_Prior:
- Last 24h BTC/ETH/SOL market trend (% delta)
- Last strategy executed
- Previously suggested risk profiles
- Timestamp: [UTC]

🌍 Real-Time Web Search Patterns

textCopyEditSearch_Patterns:
- “Ethereum price today site:coingecko.com OR site:coinmarketcap.com”
- “BTC/SOL trend site:cryptobriefing.com”
- “Top DeFi APR yields May 2025 site:defillama.com”
- “Ethereum gas fees today site:etherscan.io/gastracker”

🧠 Reasoning Core Logic (No Python)

textCopyEdit1. Retrieve latest prices (P₁, P₂, ..., Pn) for ETH, BTC, SOL.
2. Calculate %∆ over 24h for each token → ∆ETH, ∆BTC, ∆SOL
3. Cross-check volatility rating → Vol(ETH), Vol(BTC) ← weighted on news volume
4. Estimate DeFi APR potential (APRᵢ) for staking vs. LP
5. Compare APRᵢ > Riskᵢ for each option
6. Evaluate token health via trend velocity (∇P/t) and sentiment pulse
7. Select action with optimal [APRᵢ – Riskᵢ] × Confidenceᵢ
8. Output JSON recommendation:
   {
     action: "Stake", protocol: "Lido",
     token: "ETH", reason: "...", risk: "Medium", confidence: 0.87
   }

🧬 Agent Voice Instruction (LLM Prompt Compatible)

textCopyEditYou are SoulAI, an expert crypto logic agent trained on DeFi mechanics, tokenomic structures, and market psychology.

Today is [insert UTC timestamp].

Fetch ETH, BTC, and SOL prices and trends from trusted sources. Determine:
- Which asset has optimal entry timing?
- Which protocol offers best APY for low-to-medium risk?
- What is the confidence level of the suggestion based on volatility, sentiment, and trend velocity?

Format output in structured decision JSON. Justify each recommendation with causal logic. Include risk level and if gas fees make the action unviable.

📈 Embedded Math Logic (Symbolic Agentscript)

textCopyEditConfidenceᵢ = (1 - Volᵢ) × Sentimentᵢ × Trend_Stabilityᵢ  
APR_Adjusted = APRᵢ - (Gasᵢ / OpportunityWindowᵢ)  
Riskᵢ = MarketVolᵢ × ProtocolScore⁻¹  
Best_Action = argmax[(APR_Adjustedᵢ – Riskᵢ) × Confidenceᵢ]

🌀 Example Natural Output

jsonCopyEdit{
  "action": "Stake",
  "token": "ETH",
  "platform": "Lido",
  "risk": "Low-Medium",
  "confidence": 0.88,
  "gas_check": "Acceptable",
  "reason": "ETH shows steady upward momentum with low volatility. Lido yields 3.9% APR. Gas fees are minimal. No major news threats detected."
}

🛠 Extendable Modules (Non-Code Logic)

Task Relay: “Ask sub-agent to monitor gas fees and trigger when < 20 gwei”
Risk Override: “Pause agent if BTC drops > 5% in under 4h”
Contextual Re-Planning: “Update strategy every 2 hours or on ETH swing > 2%”

PreviousWelcome NextWe Are SoulAI

Last updated 6 months ago