The Agent Skills Directory

[SAFE]: The skill is a documentation and template resource for structuring AI agent interactions. It does not contain executable code, scripts, or active commands.
[DATA_EXPOSURE_AND_EXFILTRATION]: The instructions describe workflows for reading project-specific files like reference materials and source code. No access to sensitive system directories (e.g., .ssh, .aws) or credentials was found.
[COMMAND_EXECUTION]: The skill mentions placeholder variables for build, test, and lint commands (e.g., {BUILD_COMMAND}). These are standard for development agents and are intended to be configured by the user for their specific project environment.
[PROMPT_INJECTION]: The templates use standard instructional language to define agent roles and tasks. No attempts to bypass safety filters or override system constraints were detected.
[INDIRECT_PROMPT_INJECTION]: The skill outlines a pipeline that ingests external data (reference docs and existing code) into the agent's context. While this represents a potential surface for indirect injection, the skill mitigates this by providing structured templates with explicit exit criteria, task boundaries, and completion signals.

prompt-pipeline