Skip to main content
Only install skills from trusted sources—ideally ones you created yourself or from the official repository. Skills can execute bash commands, read/write files, and make network requests. A malicious skill could lead to data exfiltration, unauthorized system access, or other security risks.

​
Available Skills

Browse Skills

Browse skills in the GitHub repository

Create Your Own

Learn to build custom skills

​
Installing Skills

​
Managing Skills

​
Reload Skills

After manually creating or modifying skills: 
service: extended_openai_conversation.reload_skills
When to reload:
  • After manually creating a new skill
  • After editing SKILL.md files
  • After modifying skill metadata
You don’t need to reload after using download_skill service—it reloads automatically.

​
Security Considerations

Before installing any skill, audit it thoroughly:
  • Review all files: Check SKILL.md, scripts under scripts/, and any other bundled files. Look for unexpected network calls, unusual file access patterns, or operations that don’t match the skill’s stated purpose.
  • Be cautious with external sources: Skills that fetch data from external URLs carry additional risk—even a trustworthy skill can be compromised if its external dependencies change over time.
  • Watch for tool misuse: Skills can invoke HA services, run bash commands, and access the filesystem. Verify that a skill only uses the capabilities it claims to need.
  • Treat it like installing software: Apply the same caution you would when installing any program on your system, especially for skills that have access to sensitive data or critical automations.