The Agent Skills Directory

[SAFE]: The skill is entirely documentation-based, providing users with templates and best practices for creating AI agent skills. It contains no active scripts, remote code execution patterns, or system-modifying commands.- [METADATA_POISONING]: Analysis of the YAML frontmatter and overall metadata shows that the skill accurately describes its functionality without attempting to mislead the agent or bypass safety filters.- [EXTERNAL_DOWNLOADS]: The skill references official documentation and repositories from Anthropic (e.g., github.com/anthropics/skills), which are trusted organizations and well-known services. These references are documented neutrally and do not involve unauthorized data transfers or code execution.- [PROMPT_INJECTION]: No evidence of prompt injection, role-play bypasses, or instructions to ignore system guidelines was detected within the body or metadata of the skill.

Skill Builder