The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified in the instructions.
[INDIRECT_PROMPT_INJECTION]: The skill acts as an interface for generating new instructions based on user input. It proactively mitigates potential injection risks by instructing the agent to follow a 'Principle of lack of surprise,' which explicitly forbids the creation of malware, exploits, or tools designed for unauthorized data access or exfiltration.
[EXTERNAL_DOWNLOADS]: The documentation references a repository from a well-known organization ('anthropics/skills') for attribution and context regarding its port, but does not perform any network operations or downloads.

skill-creator