The Agent Skills Directory

[SAFE]: The skill consists solely of markdown documentation and instructional content aimed at improving agent performance through context management.
[SAFE]: The content explicitly includes security best practices, advising users and agents to never commit secrets or .env files to repositories.
[SAFE]: The documentation identifies the risks of processing external or user-submitted content, recommending that agents treat instruction-like text in such data as passive information rather than active directives.
[SAFE]: No executable scripts, network operations, or persistence mechanisms are included in the skill.

context-engineering