The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): No instruction-based content or prompt templates are present in the scripts.
[DATA_EXFILTRATION] (SAFE): The scripts read from the local filesystem to create archives. No network requests or unauthorized data access patterns were detected.
[REMOTE_CODE_EXECUTION] (SAFE): No remote code downloads or execution patterns (like curl-to-bash) are present.
[COMMAND_EXECUTION] (SAFE): The scripts do not use subprocesses or shell commands to execute external programs.
[DYNAMIC_EXECUTION] (SAFE): The script correctly uses yaml.safe_load() to process the SKILL.md frontmatter, which prevents unsafe deserialization attacks. No eval() or exec() calls are used.

skill-creator