The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Python subprocess module to interface with external command-line tools (Codex and Grok) to perform its core image generation functions. These calls are directed at established, user-installed CLI tools and are used to invoke specific image generation features.
[EXTERNAL_DOWNLOADS]: The skill includes a shell script (repair-hermes-assets.sh) to repair corrupted binary assets by downloading them from the author's GitHub repository. Additionally, the packs command in the main script allows downloading community-contributed character packs from a designated GitHub repository. These downloads target the vendor's own infrastructure and well-known services.
[INDIRECT_PROMPT_INJECTION]: The skill's primary workflow involves analyzing user-provided articles or URLs to identify 'load-bearing moments' for illustration. This ingestion of untrusted external content represents a surface for indirect prompt injection. While the skill employs prompt templates and thesis-locking steps to mitigate this, malicious instructions embedded in the processed source material could attempt to influence the generated output or agent behavior.

illo