The Agent Skills Directory

[NO_CODE]: The skill is comprised solely of Markdown documentation files including README.md, SKILL.md, and research files in the references directory. It lacks any executable code, shell scripts, or binary files that would execute on the host system or perform automation tasks.\n- [SAFE]: Detailed analysis of the skill's content identified no security risks. There are no attempts to access sensitive file paths, hardcode credentials, perform network operations, or escalate privileges. The functionality is limited to character emulation and information retrieval from its own reference files.\n- [PROMPT_INJECTION]: The role-playing instructions provided for the agent are standard for persona adoption and do not include malicious patterns such as commands to ignore safety guidelines, bypass system constraints, or reveal underlying system prompts. The skill includes explicit honesty boundaries to prevent the AI from making unfounded claims in areas like commercial or legal advice.

tong-jincheng-perspective