AI Chat Interface
The heart of Oracled is its intelligent chat interface, powered by our proprietary AI model trained on 72,000+ cryptocurrency tokens. This guide explains how the chat interface works and how to get the most out of your conversations with the AI oracle.
Overview
The AI chat interface is designed for natural, conversational interaction. You don't need to learn special commands or syntax - just ask questions in plain English, and the AI will understand and respond with detailed, data-driven insights.
Interface Components
Main Chat Area
The central chat area displays:
Your Messages: Shown on the right with a distinct background
AI Responses: Streamed in real-time on the left with markdown formatting
Timestamps: Track when messages were sent
Loading Indicators: "Oracled is thinking..." while processing
Message Input
Located at the bottom of the interface:
Text Input Field: Type your questions and requests
Token Contract Address Field (Optional): Enter contract addresses for specific token analysis
Send Button: Submit your message (or press Enter)
Cooldown Timer: Shows remaining wait time between messages
Conversation Starters
Quick-start buttons for common queries:
"What are the current trending memecoins?"
"How do I identify a potential rugpull?"
"Explain pump.fun token graduation"
"What metrics indicate a healthy token?"
Click any starter to instantly send that query to the AI.
How the AI Works
Natural Language Understanding
The AI processes your queries through advanced natural language processing:
Context Awareness
Remembers previous messages in the conversation
Understands follow-up questions without needing repetition
Maintains context about tokens you're discussing
Adapts to your level of expertise
Intent Recognition
Identifies what type of information you're seeking
Determines if you want risk assessment, price analysis, or general info
Recognizes when you need more detailed vs. summary information
Understands implicit questions and assumptions
Terminology Recognition
Knows crypto-specific terms (DYOR, FOMO, FUD, etc.)
Understands blockchain concepts (smart contracts, liquidity, etc.)
Recognizes platform names (Pump.fun, DexScreener, etc.)
Interprets contract addresses and token tickers
Real-Time Data Integration
The AI doesn't just rely on training data:
Live Data Sources
DexScreener API: Current prices, volumes, and trading data
Blockchain Explorers: On-chain transactions and holder information
Web Search: Latest news, developments, and social sentiment
Market Aggregators: Multi-source price and liquidity data
Dynamic Analysis
Combines historical patterns with current market conditions
Updates risk assessments based on real-time data
Factors in recent similar token performances
Adjusts recommendations as new information emerges
Response Generation
The AI crafts comprehensive responses:
Structured Formatting
Headers organize sections logically
Bullet points for easy scanning
Tables for comparative data
Code blocks for contract addresses
Bold text highlights key information
Evidence-Based Insights
Cites data sources when possible
Explains reasoning behind assessments
Provides specific metrics and numbers
References comparable tokens or patterns
Chat Features
Streaming Responses
Responses appear in real-time as the AI generates them:
Words stream progressively, not all at once
You can start reading while the AI continues writing
Typical streaming speed: 20-50 words per second
Full responses typically complete within 10-20 seconds
Markdown Support
The AI uses rich formatting for better readability:
Headers
## Risk Assessment
### Key FindingsLists
Bullet points for features
Numbered lists for steps
First step
Second step
Emphasis
Bold for important information
Italic for emphasis
Code blocksfor addresses
Tables
Price
$0.001
🟢 Stable
Links DexScreener - Clickable references
Token-Specific Analysis
When you provide a contract address:
Automatic Detection
AI recognizes the address format
Fetches data from multiple sources
Analyzes on-chain metrics
Generates comprehensive report
Analysis Components
Current price and market cap
24-hour trading volume
Holder count and distribution
Top holder percentages
Liquidity pool information
Contract security analysis
Risk score and recommendation
Conversation Context
The AI maintains context throughout your session:
Memory Scope
Remembers all messages in the current session
References previous tokens discussed
Builds on earlier explanations
Connects related concepts
Follow-Up Questions
User: "Tell me about [Token A]"
AI: [Provides analysis]
User: "How does it compare to [Token B]?"
AI: [Compares both, remembering Token A details]Context Reset
Refresh the page to start a new conversation
Context doesn't carry across sessions
No conversation history stored on servers
Query Types
General Questions
Ask broad questions about cryptocurrency:
"What makes a good memecoin?"
"Explain how Pump.fun works"
"What are the risks of low-cap tokens?"
"How do I evaluate a new project?"
The AI provides educational, comprehensive answers.
Token Analysis Requests
Request specific token evaluation:
"Analyze this token: [contract address]"
"Is [token name] a good investment?"
"What are the risks of [contract address]?"
"Compare these two tokens: [address 1] and [address 2]"
The AI fetches real-time data and provides detailed analysis.
Market Insights
Get current market information:
"What are today's trending tokens?"
"Show me the best performing memecoins this week"
"What's happening in the Solana ecosystem?"
"Are we in a bull or bear market?"
The AI searches for current information and summarizes trends.
Risk Assessment
Request security evaluation:
"Is this token a rugpull?"
"Check the contract security: [address]"
"What red flags should I look for?"
"Evaluate the holder distribution"
The AI analyzes security indicators and provides risk scores.
Best Practices
Asking Effective Questions
Be Specific
❌ "Tell me about tokens"
✅ "Analyze the token at address [contract address] for rugpull risks"
Provide Context
❌ "Is it good?"
✅ "Is this token good for a short-term trade or long-term hold?"
Include Contract Addresses
Use the dedicated contract address field
Or paste directly in your message
Multiple addresses can be analyzed together
Getting Better Answers
Ask Follow-Up Questions
User: "Analyze [token]"
AI: [Provides analysis with risk score: MEDIUM]
User: "What would need to change for it to be LOW risk?"
AI: [Explains specific improvements needed]Request Clarification
If response is too technical, ask for simpler explanation
Request more details on specific points
Ask for examples or comparisons
Use Natural Language
No need for formal commands
Conversational tone works best
Ask as you would ask a knowledgeable friend
Understanding Limitations
The AI is powerful but has limits:
Cannot Predict Prices
Won't give price targets or predictions
Can assess risk but not guarantee outcomes
Market is inherently unpredictable
Relies on Available Data
Analysis quality depends on data availability
Very new tokens may have limited information
Some chains have less data than others
Not Financial Advice
Provides analysis and insights only
Always do your own research (DYOR)
Never invest more than you can afford to lose
Rate Limiting
The 50-Second Cooldown
After sending a message:
Countdown Timer: Displays remaining time
Disabled Input: Fields locked during cooldown
Progress Indicator: Visual progress bar
Persistent: Survives page refreshes
Why Rate Limiting?
Necessary for:
Fair Access: Ensures all users can use the platform
System Stability: Prevents overload during high traffic
Quality Responses: AI has time to perform thorough analysis
Abuse Prevention: Stops automated bots and spam
Working Within Limits
Maximize each query:
Ask Comprehensive Questions: Include all details in one message
Multiple Topics: Can ask about several things at once
Batch Analysis: Include multiple contract addresses
Follow-Ups: Plan your next question while waiting
Privacy & Security
Data Handling
Your conversations are private:
No Storage: Messages not saved on servers
No Logging: Conversation history not tracked
Session Only: Context exists only during your session
Anonymous: No user identification or tracking
Safe Usage
Protect yourself:
Never Share: Don't share private keys or seed phrases
Verify Addresses: Double-check contract addresses
Cross-Reference: Confirm AI findings with other sources
Stay Skeptical: Don't trust blindly, always DYOR
Tips for Power Users
Advanced Queries
Get more sophisticated analysis:
Comparative Analysis
"Compare these three tokens by market cap, holder distribution,
and liquidity: [address1], [address2], [address3]"Conditional Scenarios
"If this token graduates from Pump.fun, what would be the
expected market cap range based on similar tokens?"Historical Context
"How does this token's launch pattern compare to successful
memecoins from Q4 2024?"Efficient Communication
Streamline your workflow:
Structured Requests
"For [contract address], provide:
1. Risk score with explanation
2. Top 3 red flags
3. Comparison to [similar token]
4. Short-term outlook"Batch Processing
"Quick risk assessment for these 5 tokens:
[addr1], [addr2], [addr3], [addr4], [addr5]"Troubleshooting
AI Not Responding
If the AI seems stuck:
Wait 30 seconds for processing
Check your internet connection
Refresh the page and try again
Simplify your query
Incomplete or Generic Responses
If responses lack detail:
Provide more context in your question
Include contract addresses when relevant
Ask follow-up questions for clarification
Specify what type of information you need
Rate Limit Issues
If cooldown seems incorrect:
Check browser local storage
Try incognito/private mode
Wait for the timer to expire naturally
Don't refresh repeatedly (doesn't help)
Next Steps
Master Token Analysis: Learn how to analyze tokens effectively
Understand Risk Scores: Read about risk score interpretation
Best Practices: Check out tips and best practices
AI Training: Learn about our AI model
Ready to start chatting? Visit oracled.xyz and ask your first question!
Last updated