Goals
Goals
Cod3x agents are built around your trading goals — natural-language statements that define what you want to achieve and how your agent should act to get there, all automated.
A goal is more than a single trade or setup. It’s a living instruction that guides how your agent thinks, reacts, and executes across changing market conditions.
Examples:
“Accumulate ETH on dips while maintaining 2x leverage.”
“Scalp high-volume altcoins using RSI and VWAP.”
“Exit all positions if BTC breaks key support.”
Each goal becomes a self-contained workflow — your agent analyzes the market, creates or updates positions, manages risk, and reports results automatically. You can set or adjust goals at any time through chat or manually.
Click on any goal you have to dive in deeper and see it's details or generate a whole new template agent based on the ones you've created.
Creating a New Goal
Each goal defines what your Cod3x agent should accomplish — from research and analysis to execution and reporting. Goals are modular, autonomous routines that can run on schedule or be triggered on-demand, giving your agent direction and focus.
Step 1: Basic Information
In this step, you’ll define the purpose and priority of your goal.
Goal Title — Give your goal a clear name (e.g., “Scan top altcoin markets for long setups”).
Description — Describe what the goal should achieve in plain language. Example: “Analyze trending tokens on GMX and identify 3 high-confluence long setups using RSI and MACD.” (You can see more examples on how detailed goals descriptions can look like in some of the templated agent's goals)
Type — Select the goal category:
Research – For analysis, data gathering, or sentiment scanning.
Trading – For placing trades or managing open positions.
Priority — Define urgency from Low (1) to Urgent (5). Higher priority goals will appear first in your agent’s task queue and influence execution order.
Once you’ve filled these out, click Next to proceed to configuration.
Step 2: Configuration
Here you’ll decide how your goal will run and which model powers it.
AI Model Selection
Cod3x supports multiple large language models (LLMs), each optimized for different workloads. Choose the one that best fits your goal’s complexity and budget.
Anthropic: Claude Sonnet 4.5 – Deep reasoning, high context, recommended for advanced goal logic.
OpenAI: GPT-4.1 – Balanced performance for analysis and execution tasks.
GPT-4.1 Mini – Fast and cost-efficient for frequent, lighter-weight goals.
xAI: Grok 3 – High-context, creative reasoning useful for adaptive market reads.
Once selected, you’ll see:
Context Window – The total token capacity available (e.g., 1000k tokens).
Cost Breakdown – Estimated compute usage and credits per execution.
Provider Details – Model source (e.g., OpenRouter).
You can view the full list of models anytime under View All Models.
Toggle the "Requires Open Position" toggle if this is a goal that reviews current open positions.
Tools & Capabilities
Optional: Attach any tools your goal will use — such as on-chain data fetchers, price analysis modules, or order executors. By default the agent dynamicallys pick the best tool for each run.
Click Next to move to scheduling.
Step 3: Schedule
This step defines when and how often the goal runs.
Schedule Type – Choose between:
Recurring – Runs automatically on selected days and times.
Manual – Executes only when triggered by you or another goal.
Daily – Executes daily on selected execution times
Days of Week – Select which days to execute.
Execution Times – Add one or more run times (e.g., 21:35).
Timezone – Set your preferred timezone for consistency across agents.
You can add multiple schedules to align analysis and execution goals (for example, run “Market Scan” every morning before your “Trade Execution” goal).
Once saved, your goal becomes active and visible in the dashboard. Your agent will now monitor conditions and execute as defined.
Last updated