The Agent Skills Directory

[SAFE]: The skill performs local file read and write operations within a project-specific directory structure (.adlc). It follows a "least privilege" approach by reading only matched context modules and explicitly forbidding directory traversal. No network operations or external dependencies are involved.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from local configuration files (.adlc/init-options.json) and documentation modules (rules, personas, examples). While this constitutes an attack surface, the risk is negligible as the data source is the user's own project environment, and the output is structured into summary tables with specific constraints on fabrication.

levelup-boot