oracle

To provide AI models with code context for debugging, you bundle prompts with selected source files and error messages, allowing the AI to generate detailed analysis and potential fixes based on that specific code.

The best way to use an AI assistant for code analysis is to feed it specific source files and task prompts together, ensuring the model receives the exact context needed to generate accurate advisory output and recommendations.

Yes, you can use both API and browser engines for code analysis tasks. For long-running tasks, the process requires specific engine configuration and session management to maintain context and deliver advisory output.

You can preview the token count for a task by selecting files in your source directory and using the skill to evaluate the bundled prompts and code context before submitting them to the AI model.

Yes, specific file selection is required. You must manually choose the relevant source files and error messages to bundle with your prompt so the AI model receives the precise code context needed for accurate advisory output.

Prompt engineering requires bundling files with code context because AI models need relevant source code and error details directly embedded in the prompt to generate accurate debugging analysis and task completion recommendations.