Skip to content

CodeAudit Decorator

Requests an audit of existing code to identify issues before making changes.

Category: Systematic Debugging

Parameters

Parameter Type Description Default
scope enum Audit scope component
focus enum Audit focus areas all
output enum Audit output format detailed

Scope Options

  • function: Focus your audit on the specific function provided.
  • component: Examine the entire component for issues.
  • module: Conduct a thorough audit of the complete module.
  • system: Perform a system-wide audit to identify issues across all components.
  • specific-issue: Focus your audit specifically on the issue described.

Focus Options

  • bugs: Concentrate on identifying logical errors, edge cases, and potential runtime exceptions.
  • performance: Analyze for performance bottlenecks, inefficient algorithms, and resource usage issues.
  • security: Identify security vulnerabilities, injection risks, and data protection concerns.
  • maintainability: Evaluate code quality, documentation, and adherence to best practices.
  • all: Examine all aspects including bugs, performance, security, and maintainability.

Output Options

  • summary: Provide a concise summary of the key issues found.
  • detailed: Deliver a comprehensive report with detailed explanations of each issue.
  • categorized: Organize findings by category (bugs, performance, security, etc.).
  • prioritized: Present issues in order of importance, with critical issues first.

Examples

Basic code audit for a payment module

+++CodeAudit(scope=module, focus=bugs, output=prioritized)
Audit the payment processing module to identify potential bugs and issues. Provide a prioritized list of problems found.

The model will perform a thorough audit of the payment processing module, focusing specifically on bugs and logical errors, and will present the findings as a prioritized list with the most critical issues first.

Security-focused audit of a component

+++CodeAudit(scope=component, focus=security, output=detailed)
Review this authentication component for security vulnerabilities.

The model will examine the authentication component with a focus on security vulnerabilities, providing a detailed report of all potential security issues found.

Model-Specific Implementations

gpt-4-turbo

Instruction: Analyze this code carefully before suggesting any changes. Look for issues related to {focus} within the {scope} and provide a {output} report.

Notes: Simplified instruction for models with more limited context windows.

Implementation Guidance

Code review for a payment processing system

Original Prompt:

Refactor this payment processing code to improve performance.

Transformed Prompt:

Before making any changes, please perform a code audit to identify potential issues. Conduct a thorough audit of the complete module. Analyze for performance bottlenecks, inefficient algorithms, and resource usage issues. Deliver a comprehensive report with detailed explanations of each issue.

Refactor this payment processing code to improve performance.

Notes: The decorator adds instructions to first audit the code before proceeding with the requested refactoring.

Transformation Details

Base Instruction: Before making any changes, please perform a code audit to identify potential issues.

Placement: prepend

Composition Behavior: accumulate

Parameter Effects:

  • scope:
  • When set to function: Focus your audit on the specific function provided.
  • When set to component: Examine the entire component for issues.
  • When set to module: Conduct a thorough audit of the complete module.
  • When set to system: Perform a system-wide audit to identify issues across all components.
  • When set to specific-issue: Focus your audit specifically on the issue described.

  • focus:

  • When set to bugs: Concentrate on identifying logical errors, edge cases, and potential runtime exceptions.
  • When set to performance: Analyze for performance bottlenecks, inefficient algorithms, and resource usage issues.
  • When set to security: Identify security vulnerabilities, injection risks, and data protection concerns.
  • When set to maintainability: Evaluate code quality, documentation, and adherence to best practices.
  • When set to all: Examine all aspects including bugs, performance, security, and maintainability.

  • output:

  • When set to summary: Provide a concise summary of the key issues found.
  • When set to detailed: Deliver a comprehensive report with detailed explanations of each issue.
  • When set to categorized: Organize findings by category (bugs, performance, security, etc.).
  • When set to prioritized: Present issues in order of importance, with critical issues first.

Compatibility

  • Requires: None
  • Conflicts: QuickFix
  • Compatible Models: gpt-4o, gpt-4-turbo, claude-3-7-sonnet-latest, llama-3.2
  • Standard Version: 1.0.0 - 2.0.0
  • CodeRefactor: Enhances CodeAudit CodeAudit works well before CodeRefactor to ensure issues are identified before making changes.
  • QuickFix: Conflicts with CodeAudit CodeAudit's thorough analysis approach conflicts with QuickFix's rapid solution focus.