The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary function involves executing shell commands to drive applications (servers, UIs, and CLIs) during the Live pass phase (Step 4), based on the launch model of the target verification skill.
[INDIRECT_PROMPT_INJECTION]: The skill reads project source code and feature files (Step 2) to generate live-verification recipes, creating a surface for indirect prompt injection where malicious code comments or instructions could influence the agent's actions during the Live pass.
Ingestion points: Project source code and feature map files located in the target verification skill's directory.
Boundary markers: There are no specific instructions or delimiters used to ensure the agent ignores potentially malicious instructions embedded within the source code being analyzed.
Capability inventory: The skill has the capability to execute shell commands, launch sub-agents, and modify or create executable harness scripts.
Sanitization: The skill lacks validation or sanitization mechanisms for the recipes derived from source code analysis before they are executed.
[DYNAMIC_EXECUTION]: In the Triage phase (Step 5), the agent is instructed to create or modify executable harness scripts to address gaps. This dynamic generation and execution of code based on untrusted project data is a potential risk vector.

maintain-verification-skill