The Agent Skills Directory

[SAFE]: The skill implements asset management through predefined API endpoints (/api/tools/library/search and /api/tools/library/apply), which limits the scope of file operations to legitimate registry tasks.
[SAFE]: Secure credential management is observed through the use of an environment-injected OD_TOOL_TOKEN, avoiding hardcoded secrets or unsafe token handling.
[SAFE]: No indicators of remote code execution, prompt injection, or data exfiltration are present in the skill instructions or tool definitions.
[SAFE]: The skill operates within its stated purpose of curating library assets without requesting excessive permissions or performing suspicious network activities.

library-curator