The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands, including directory checks using ls and running Python scripts such as validate_contrast.py and lint_hardcodes.py. These scripts are located in relative paths (../../../scripts/) outside the skill's specific folder.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection due to its core functionality of processing external data.
Ingestion points: The skill is designed to scan the user's codebase (Step 2) and read various local markdown files for workflows and design rules (Steps 1, 3, 6).
Boundary markers: The instructions lack explicit delimiters or warnings to the agent to ignore any natural language instructions that might be embedded within the scanned codebase or configuration files.
Capability inventory: The skill has the capability to execute shell commands and run Python scripts on the host system.
Sanitization: There is no evidence of sanitization, filtering, or escaping of the scanned content before it is processed by the agent's logic.

redesign