The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands via mkdir -p to prepare the output directory structure for audit findings. This is a standard and necessary operation for an automated reporting tool.
[PROMPT_INJECTION]: The instructions include directives for agent autonomy and lifecycle management, such as 'shutdown_request' and 'shutdown_response' protocols. These are used for orchestration between the team leader and worker agents and do not attempt to bypass core safety constraints.
[DATA_EXPOSURE]: The skill operates on a user-defined source_path and output_path. It identifies and maps sensitive areas like routes and authentication logic within the source code for auditing purposes, which is the primary function of the skill.
[REMOTE_CODE_EXECUTION]: While the skill orchestrates other analysis tools (e.g., /java-route-mapper, /java-sql-audit), these are referenced as internal skill components within the environment rather than external remote scripts.

java-audit-pipeline