The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from a user-specified project path, which could lead to instructions in those files influencing the agent.
Ingestion points: The scripts/api_validator.py script accepts a <project_path> as an argument in SKILL.md, indicating it reads and analyzes external files.
Boundary markers: There are no instructions defining delimiters or safety warnings for the agent when processing the results of this analysis.
Capability inventory: The skill executes a Python script (scripts/api_validator.py) to analyze project files.
Sanitization: No sanitization or validation of the ingested content is described in the skill instructions.
[COMMAND_EXECUTION]: The skill defines a command for executing a local script to perform its primary validation task.
Evidence: The script table in SKILL.md specifies the command python scripts/api_validator.py <project_path>.

Api Patterns