The Agent Skills Directory

[SAFE]: The skill contains natural language instructions that guide the agent to reformat its output. It does not attempt to bypass safety guardrails or extract sensitive information.
[SAFE]: No network operations, external downloads, or file system modifications are performed. The reference to CONTEXT.md is a local file reference for terminology consistency.
[SAFE]: The skill configuration utilizes security best practices by setting disable-model-invocation: true and allow_implicit_invocation: false, limiting the skill's capabilities to prompt-based instruction following only.

wait-what