> ## Documentation Index > Fetch the complete documentation index at: https://docs.gaife.com/llms.txt > Use this file to discover all available pages before exploring further. # Coder Task > This page gives you an overview of the Coder Task in Agent Builder ## Overview Coder Tasks allow you to execute pre-defined code artifacts within your workflow. These tasks integrate the computational power of [Coder Artifacts](../knowledge/coder-artifact.md) into your workflow processes. Coder Tasks are ideal for custom calculations, data transformations, and complex business logic implementation. ## Visual Example Coder Task Example ## Configuration Structure ```json theme={null} { "type": "CODER", "block": { "name": "Coder Task Name", "type": "CODER", "instructions": "Task instructions", "code_artifact_id": 123, "input_parameters": [], "expected_output": [], "dependencies": [], "error_policy": "RAISE" } } ``` ## Required Fields | Field | Type | Description | Required | | :----------------- | :------ | :------------------------- | :------: | | name | string | Task identifier | ✅ | | instructions | string | Task instructions | ✅ | | code\_artifact\_id | integer | Reference to code artifact | ✅ | | input\_parameters | array | Input configuration | ✅ | | expected\_output | array | Output configuration | ✅ | | dependencies | array | Task dependencies | ❌ | | error\_policy | string | Error handling strategy | ❌ | ## Input Parameters The input parameters define what data your Coder Task will receive. These parameters will be passed to the `arguments` dictionary in your [Coder Artifact](../knowledge/coder-artifact.md) function. ### Basic Input ```json theme={null} { "name": "data", "type": "STRING", "description": "Input data for processing", "required": true, "source": "task_config" } ``` ### Complex Input ```json theme={null} { "name": "config", "type": "OBJECT", "description": "Configuration object", "properties": [ { "name": "algorithm", "type": "STRING", "description": "Algorithm selection", "required": true }, { "name": "parameters", "type": "OBJECT", "description": "Algorithm parameters", "properties": [ { "name": "iterations", "type": "INTEGER", "description": "Number of iterations", "required": true } ] } ] } ``` ## Output Parameters Output parameters define the structure of data that your Coder Task will return. These should match the structure returned by your [Coder Artifact](../knowledge/coder-artifact.md) function. ### Simple Output ```json theme={null} { "name": "result", "type": "STRING", "description": "Processing result" } ``` ### Complex Output ```json theme={null} { "name": "analysis_result", "type": "OBJECT", "properties": { "result": { "type": "FLOAT", "description": "Calculated value" }, "metadata": { "type": "OBJECT", "description": "Processing metadata" }, "logs": { "type": "ARRAY", "items": { "type": "STRING" }, "description": "Processing logs" } } } ``` ## Workflow Integration Coder Tasks act as computational nodes in your workflow, taking input from previous tasks, processing it using the referenced Coder Artifact, and outputting results for subsequent tasks. ## Common Use Cases ### 1. Data Transformation ```json theme={null} { "name": "Transform Data", "code_artifact_id": 123, "instructions": "Transform input data according to specified rules", "input_parameters": [ { "name": "data", "type": "OBJECT", "description": "Raw data object", "required": true }, { "name": "transformation_rules", "type": "ARRAY", "description": "Transformation rules", "required": true } ], "expected_output": [ { "name": "transformed_data", "type": "OBJECT", "description": "Transformed data" } ] } ``` ### 2. Custom Calculation ```json theme={null} { "name": "Calculate Metrics", "code_artifact_id": 456, "instructions": "Calculate custom business metrics", "input_parameters": [ { "name": "raw_metrics", "type": "ARRAY", "description": "Raw metric data", "required": true } ], "expected_output": [ { "name": "calculated_metrics", "type": "OBJECT", "properties": { "kpi1": "FLOAT", "kpi2": "FLOAT", "summary": "STRING" } } ] } ``` ### 3. Invoice Processing ```json theme={null} { "name": "Calculate Invoice Information", "code_artifact_id": 789, "instructions": "Compute the total invoice amount based on the collected item details", "input_parameters": [ { "name": "items_to_be_billed", "type": "ARRAY", "items": { "type": "OBJECT", "properties": { "item_price": { "type": "STRING" }, "item_quantity": { "type": "STRING" } } }, "description": "List of items with price and quantity", "required": true } ], "expected_output": [ { "name": "calculation_amount", "type": "OBJECT", "properties": { "total_tax": { "type": "STRING" }, "current_date": { "type": "STRING" }, "total_amount": { "type": "STRING" }, "total_state_tax": { "type": "STRING" }, "total_central_tax": { "type": "STRING" }, "total_amount_with_gst": { "type": "STRING" } }, "description": "Calculated invoice amounts" } ] } ``` ## Best Practices ### 1. Code Artifact Management ✅ **Do**: * Use version control for your Coder Artifacts * Document dependencies clearly * Include test cases with your Coder Artifacts * Handle errors gracefully in your Python code (as shown in the [Coder Artifact documentation](../knowledge/coder-artifact.md)) ❌ **Don't**: * Hard-code configurations that might change * Ignore error cases in your Coder Artifact code * Skip input validation in your Python function * Mix business logic between Coder Tasks and Coder Artifacts ### 2. Input Parameters ✅ **Do**: * Validate all inputs in your Coder Artifact code * Document input requirements thoroughly * Use appropriate types that match your Coder Artifact expectations * Include examples in your documentation ❌ **Don't**: * Skip validation in your Coder Artifact code * Use ambiguous parameter names * Ignore data types that will cause conversion issues ### 3. Output Handling ✅ **Do**: * Define clear output structure that matches your Coder Artifact return values * Include error information in your response * Add processing metadata when useful * Validate outputs in your Coder Artifact code ❌ **Don't**: * Return raw errors without context * Skip validation of output data * Use inconsistent formats between Coder Artifact and Coder Task configurations ## ⚠️ Error Handling ### Configuration ```json theme={null} { "error_policy": "RAISE", "retry_count": 3, "retry_delay": 60 } ``` ### Error Types 1. **Input Validation Errors**: Occurs when input doesn't match expected format 2. **Processing Errors**: Occurs during execution of the Coder Artifact 3. **System Errors**: Occurs due to system limitations or failures ## 🔗 Integration Examples ### With AI Task ```mermaid theme={null} graph TD A[AI Task] --> B[Coder Task] B --> C[App Task] ``` ### With Data Processing ```mermaid theme={null} graph TD A[Data Input] --> B[Coder Task] B --> C[Data Validation] C --> D[Result Output] ``` ## ❓ Common Issues and Solutions | Issue | Solution | | :---------------------- | :--------------------------------------------------------------- | | Code Artifact Not Found | Verify artifact ID and ensure it's published | | Input Validation Failed | Check input types and format against Coder Artifact requirements | | Processing Error | Review error logs in your Coder Artifact and retry | | Memory Issues | Optimize data handling in your Python code | ## 🏢 Examples by Industry ### Finance ```json theme={null} { "name": "Risk Calculator", "code_artifact_id": 789, "instructions": "Calculate investment risk metrics", "input_parameters": [ { "name": "portfolio_data", "type": "OBJECT", "description": "Portfolio information", "required": true } ], "expected_output": [ { "name": "risk_metrics", "type": "OBJECT", "properties": { "var": "FLOAT", "sharpe_ratio": "FLOAT", "risk_level": "STRING" } } ] } ``` ## 🔄 Connection with Coder Artifacts Coder Tasks rely on well-designed Coder Artifacts to function properly. When creating a Coder Task: 1. **Select an existing Coder Artifact** from your knowledge base 2. **Configure input parameters** that match the expectations of your Coder Artifact's function 3. **Define expected output** that aligns with what your Coder Artifact function returns 4. **Provide clear instructions** for maintainers to understand the task's purpose For detailed information on creating effective Coder Artifacts, refer to the [Coder Artifact documentation](../knowledge/coder-artifact.md).