context-engineering

Manage AI agent context by creating rule files for project conventions and loading relevant specification or source files to curate the context hierarchy and improve output accuracy.

AI agent output degrades when session context becomes cluttered or irrelevant. You can resolve this by refreshing the session context and loading specific rule files to restore focus and accuracy.

Context hierarchy in AI agent session management is the structured configuration of rule files and source specifications that dictates how the agent prioritizes information to maintain focus and output quality.

Set up context for a new coding session by running the context-engineering skill with appropriate rule files to define constraints, then loading the specific source files and specifications required for your current task.

Yes, you can use rule files to enforce project-wide conventions on an AI agent by defining constraints and specifications that the agent must follow during session management and code generation.

Troubleshoot AI agent errors by providing error outputs and test results into the curated context, which guides the agent through debugging by loading relevant specifications and source files for the specific failure.