How Agentical Works
Discover the complete trading workflow from token discovery to execution and the multi-agent system behind it.
Overview
Agentical operates through a sophisticated multi-stage process that transforms raw market data into strategic trading decisions. From the moment you activate the agent, a complex workflow begins—continuously scanning the Solana blockchain, analyzing hundreds of data points, evaluating risks, and executing trades with precision timing.
This page takes you through the complete journey: from initial token discovery on pump.fun to final trade execution and position management. Understanding this process will help you appreciate the complexity working behind the scenes and make informed decisions about configuring your trading strategy.
Stage 1: Market Scanning & Token Discovery
Continuous Market Monitoring
The moment you activate Agentical, scanner agents begin their work. These specialized agents connect directly to pump.fun's infrastructure, monitoring every new token launch in real-time.
What Scanner Agents Track:
🆕 New token launches on pump.fun
📈 Migration events to DEX platforms
💹 Sudden volume spikes in existing tokens
💧 Liquidity pool formations and changes
🔄 Token age and lifecycle stage
Scanner agents don't just passively observe—they actively filter and prioritize. Tokens are immediately categorized based on their lifecycle stage, with particular attention paid to those approaching or completing migration to DEX Screener.
Initial Filtering Criteria
Before deeper analysis begins, tokens must pass basic eligibility requirements:
Token Age
Typically 2+ hours old
Avoid brand new, untested launches
Minimum Holders
User-defined threshold
Ensure basic community interest
Liquidity
Minimum pool depth
Enable actual trading capability
Migration Status
Approaching or completed
Focus on validated tokens
Tokens that fail these initial filters are discarded immediately, allowing the system to focus computational resources on genuine opportunities.
Stage 2: Multi-Factor Analysis
Comprehensive Token Evaluation
Tokens that pass initial filtering enter the comprehensive analysis phase. Multiple specialized agents work in parallel, each examining different aspects of the token's profile.
Social Media Analysis Agent
This agent evaluates the token's social presence and community legitimacy:
Twitter/X Evaluation:
Account age and creation date
Follower count and growth pattern
Engagement rate (likes, retweets, comments)
Blue checkmark verification status
Posting frequency and consistency
Community interaction quality
Telegram Assessment:
Group member count
Message frequency and activity
Bot-to-human ratio analysis
Admin responsiveness
Community sentiment
A token might have 50,000 Twitter followers, but if the engagement rate is suspiciously low or the account was created days before the token launch, red flags are raised.
Holder Distribution Analysis Agent
This agent examines on-chain holder data to detect concentration risks and manipulation patterns:
Key Metrics Analyzed:
Total unique holder count
Top 10 holder concentration percentage
Suspicious wallet clustering (connected wallets)
Dev wallet holdings and lock status
Recent large holder movements
Chart Pattern Recognition Agent
Technical analysis agents evaluate price action and trading patterns:
Trend Analysis - Uptrend, downtrend, or consolidation identification
Volume Patterns - Healthy volume distribution vs. manipulation spikes
Support/Resistance - Key price levels that might influence entries/exits
Candlestick Patterns - Bullish or bearish formation recognition
All-Time High Analysis - Distance from ATH and recovery potential
Pattern Evaluation Matrix:
📈 Healthy uptrend with consolidation
Priority consideration
🚀 Parabolic pump without pullback
Wait for consolidation
📉 Downtrend after ATH
Avoid completely
🔄 Range-bound accumulation
Monitor for breakout
⚡ Volume spike on green candles
Evaluate entry timing
Transaction Analysis Agent
This agent examines the buy/sell dynamics and transaction patterns:
Buy vs. Sell Ratio - Healthy buying pressure indicators
Transaction Frequency - Consistent activity vs. sporadic spikes
Average Transaction Size - Retail vs. whale activity patterns
Wallet Coordination - Detection of synchronized buying/selling
Bot Activity - Identification of automated trading patterns
Coordinated buying from multiple wallets created at the same time suggests manipulation. Genuine organic growth shows diverse transaction sizes from wallets of varying ages.
Security & Contract Analysis Agent
The security agent performs automated smart contract review:
Contract Checks:
✅ Ownership renounced or locked
✅ Liquidity locked with verified provider
✅ Mint authority disabled
✅ No hidden transfer restrictions
✅ No suspicious backdoor functions
✅ Standard token implementation
Website & Documentation Review:
Website functionality and legitimacy
SSL certificate validation
Domain age and registration info
Whitepaper/documentation availability
Project roadmap clarity
Stage 3: Risk Scoring & Decision Making
Comprehensive Risk Assessment
After multi-factor analysis completes, all data feeds into the risk assessment engine. This system assigns a numerical risk score based on weighted factors.
Risk Scoring Model:
Criteria Matching
The risk score is only part of the decision. Tokens must also match your selected trading criteria:
Example Criteria Configuration:
Minimum risk score: 65
Minimum holders: 1,000
Maximum top holder %: 5%
Required: Twitter blue check
Required: Locked liquidity
Chart pattern: Uptrend or consolidation
Token age: 4-48 hours
Only tokens that achieve the required risk score AND match all specified criteria proceed to execution.
Stage 4: Entry Point Optimization
Strategic Timing
Even when a token passes all criteria, timing the entry correctly is crucial. The execution agents don't immediately place orders—they wait for optimal entry conditions.
Entry Optimization Factors:
Price Position
Not at or near all-time high
Recent Movement
No immediate pump (avoid FOMO entries)
Volume
Steady or increasing, not declining
Buy Pressure
Positive buy/sell ratio
Chart Pattern
Consolidation or early breakout
Order Book
Sufficient liquidity for desired position size
Position Sizing
The agent automatically calculates appropriate position size based on:
Your Budget Settings - Per-trade and daily limits
Account Balance - Percentage-based allocation
Risk Score - Lower scores get smaller positions
Liquidity Available - Ensures minimal slippage
Strategy Configuration - Conservative vs. aggressive sizing
A higher-risk token (score 60-70) might receive 2-3% of your allocated capital, while a premium opportunity (score 85+) could receive 5-8%.
Stage 5: Trade Execution
Fast & Efficient Order Placement
Once the optimal entry point is identified, execution happens rapidly:
Execution Process:
Execution Speed Metrics:
Average execution time: < 2 seconds
Transaction confirmation: 1-3 blocks
Platform fee: 1.5% per transaction
Slippage tolerance: Dynamically optimized
Recording Trade Details
Every executed trade is immediately logged with complete details:
Entry price and timestamp
Token amount purchased
Total cost including fees
Risk score at entry
Criteria that triggered the trade
Expected hold duration range
Initial stop-loss and take-profit levels
This information becomes immediately visible in your dashboard, allowing real-time trade tracking.
Stage 6: Position Monitoring
Continuous Market Surveillance
The moment a trade executes, monitoring agents activate. These agents watch your open positions continuously, tracking market conditions and evaluating exit criteria.
What Monitoring Agents Track:
📊 Price Movement
Current price vs. entry price
Profit/loss percentage
Distance to stop-loss
Distance to take-profit targets
📈 Market Conditions
Volume changes
Holder count changes
New whale wallet activity
Social sentiment shifts
⚠️ Risk Indicators
Unusual sell pressure
Coordinated dumps
Liquidity removal signs
Developer wallet movements
Dynamic Position Management
Monitoring isn't passive—the agent actively manages positions based on evolving conditions:
Exit Triggers:
🎯 Take Profit
Target profit % reached
Sell position
🛡️ Stop Loss
Loss threshold exceeded
Immediate exit
⏰ Time-Based
Maximum hold duration
Review & decide
⚠️ Risk Alert
New red flags detected
Early exit
📉 Volume Drop
Significant liquidity decrease
Close position
Profit Taking Strategy
The agent can employ different exit strategies based on configuration:
Single Exit Strategy:
Set a single profit target (e.g., 50%)
Sell entire position when reached
Scaled Exit Strategy:
Sell 50% at first target (e.g., 30% profit)
Sell 30% at second target (e.g., 50% profit)
Let remaining 20% ride with trailing stop
Trailing Stop Strategy:
Follow price upward with dynamic stop-loss
Lock in profits as price increases
Exit on reversal
Your strategy configuration determines which approach the agent uses.
Stage 7: Trade Closure & Reporting
Exit Execution
When exit conditions are met, the agent executes the sell order with the same speed and precision as entry:
Immediate market sell order placement
Optimized for minimal slippage
Multiple RPC node submission
Confirmation tracking and verification
Performance Recording
Each completed trade generates a comprehensive report:
Trade Summary Includes:
Entry and exit prices with timestamps
Hold duration (hours/days)
Gross profit/loss percentage
Net profit/loss after fees
Risk score at entry
Reason for exit (target reached, stop-loss, time, manual)
Market conditions during hold period
Learning & Adaptation
Trade outcomes feed back into the system's learning mechanisms:
Successful patterns are reinforced
Failed trades trigger criteria review
Risk score accuracy is calibrated
Entry/exit timing is optimized
This continuous feedback loop ensures the agent improves over time, adapting to market evolution and refining its decision-making process.
User Control Throughout the Process
Intervention Points
While the agent operates automatically, you maintain control at every stage:
You Can:
⏸️ Pause the agent at any time (ongoing trades complete)
🛑 Stop the agent immediately (cancels pending orders)
💰 Manually close positions from your wallet
📊 Adjust criteria on the fly
💵 Change budget limits mid-session
✋ Override agent decisions
👀 View complete reasoning for every action
Dashboard Visibility
The dashboard provides real-time transparency into agent activities:
Workflow Efficiency
Parallel Processing
The agent's architecture allows simultaneous handling of multiple tasks:
This parallel architecture ensures no opportunities are missed while the agent manages existing positions.
Resource Management
The system intelligently allocates computational resources based on:
Number of open positions vs. available slots
Urgency of different tasks
Market volatility levels
User-defined priorities
During high-activity periods, the agent may focus more on monitoring open positions than scanning for new entries, ensuring existing capital is protected.
The Complete Cycle
Understanding how Agentical works reveals the sophistication behind its simple interface. From the moment you click "Start Agent," a complex orchestration begins—scanning thousands of tokens, analyzing hundreds of data points, evaluating risks through multiple lenses, timing entries with precision, and monitoring positions with unwavering attention.
Every decision is logged, every trade is documented, and every outcome feeds back into the system's improvement. While you focus on your daily life, Agentical works tirelessly, executing your strategy with mechanical precision and logical consistency.
The result is automated trading that combines the speed and discipline of AI with the strategic oversight and control that only you can provide.
Last updated