data-access-patterns

To reduce token usage during JSON session data access, route queries through dedicated scripts to retrieve specific fields, minimizing token overhead and speeding decision making. This pattern enforces scripted queries for JSON formats while reading Markdown content directly.

The best way to list open tasks efficiently is using pattern-based access scripts that query task data directly. By executing a task-list script with status parameters, you retrieve targeted task details without loading entire context files into the prompt.

Yes, you can use Markdown reads for exploration context alongside scripted JSON queries. This approach reads exploration notes and docs directly when needed, while routing structured session and task data through specialized scripts to optimize token efficiency.

To check the current session phase without loading the entire context file, execute a dedicated session-get script targeting the phase field. This retrieves only the required phase data, drastically cutting token consumption during planning and execution workflows.

No specific dependencies are required to run these data access scripts. The skill operates independently using bundled scripts and references, allowing you to query session, context, and task data directly within your automation environment.

You should use specialized scripts instead of direct file reads for context data to enforce token-efficient patterns. Scripted queries extract only necessary fields from JSON structures, preventing token bloat from loading entire files and improving AI prompt efficiency.