The Agent Skills Directory

[SAFE]: The skill provides philosophical and behavioral guidelines for the agent to manage its identity and responsibilities across different contexts. It does not contain any instructions that attempt to bypass safety filters or override system-level controls.
[NO_CODE]: The skill consists entirely of Markdown text and YAML configuration; no executable scripts, binaries, or shell commands are present.
[PROMPT_INJECTION]: While the skill defines a working attitude and persona management instructions, it explicitly directs the agent to 'Respect higher instructions, real access controls, actual capability, and physical limits,' which reinforces rather than undermines safety protocols.

self