The Agent Skills Directory

[SAFE]: The skill instructions and reference materials (remote-work-ops.md, kr-flexible-work-law.md, etc.) are strictly informational and procedural, focusing on human resources management.
[SAFE]: Integration with the mcp__sequential-thinking__sequentialthinking tool is a standard implementation for handling complex, multi-step planning tasks in AI agents.
[SAFE]: The skill's workflow includes routing to other vendor modules (e.g., moai-coworker:ai-slop-reviewer, moai-writer:korean-humanize). These are identified as legitimate vendor-owned resources for text refinement and present no security risk.
[INDIRECT_PROMPT_INJECTION]: The skill processes user queries to generate policies. While it handles untrusted user input, it does so within a safe context (text generation) without exposing sensitive capabilities or network access to the end-user data.
[SAFE]: No obfuscated URLs, base64-encoded payloads, or unauthorized data exfiltration patterns were detected across the analyzed files.

hr-operations