The Agent Skills Directory

[SAFE]: No malicious patterns, such as hardcoded credentials, unauthorized network communication, or persistent mechanisms, were detected across the skill's instructional files.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes untrusted data. 1. Ingestion points: User-supplied text enters the agent context via the arguments in the /人话, /humanize, and /detect commands defined in SKILL.md. 2. Boundary markers: No delimiters or explicit instructions to ignore embedded commands are used when interpolating this text. 3. Capability inventory: The skill consists entirely of text-based rules and has no access to subprocess calls, file-writing, or network tools. 4. Sanitization: No validation or escaping is performed on the ingested content. Despite the surface, the risk is negligible given the lack of exploitable tools.
[NO_CODE]: This skill is entirely instructional and does not include any Python scripts, Node.js packages, or executable binaries. All logic is implemented via natural language directives and pseudo-code patterns within Markdown files.

humanizer