The Agent Skills Directory

[SAFE]: The skill provides guidelines for creative writing research and logical reasoning. No malicious code, obfuscation, or data exfiltration attempts were identified during the analysis of the skill content.
[PROMPT_INJECTION]: The prompt instructions use standard role-assignment language ("You have been injected with...") common in agent skills and do not attempt to bypass safety filters or override system constraints.
[EXTERNAL_DOWNLOADS]: The skill encourages the use of a web_search tool to fetch real-world information. This is a primary feature for the intended purpose of novel research and does not involve downloading executable content.
[DATA_EXFILTRATION]: The skill references local file system paths for reading and writing project-specific data (e.g., novel/, 设定/), which is consistent with the stated purpose of managing creative writing assets and does not involve unauthorized access to sensitive system files.

novel-research