The Agent Skills Directory

[SAFE]: Analysis of the skill files, instructions, and scripts shows no evidence of malicious behavior, credential harvesting, or unauthorized network operations. The skill is correctly structured for its intended educational and utility purposes.
[COMMAND_EXECUTION]: The skill includes a Python utility (scripts/scaffold.py) designed to generate boilerplate code templates. This script uses standard libraries and performs local file writing operations that are appropriate for its function and do not introduce security risks.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface through its code review feature.
Ingestion points: The agent ingests user-provided code for review and refactoring suggestions as described in SKILL.md.
Boundary markers: There are no specific delimiters defined in the instructions to isolate user-provided code from the system prompt.
Capability inventory: The skill includes file system access (write) through the scripts/scaffold.py script.
Sanitization: The skill does not implement specific input sanitization for reviewed code, relying on the agent's built-in safety mechanisms.
This surface is standard for AI-based development assistants and does not present an elevated risk given the skill's capabilities.

design-patterns