The Agent Skills Directory

[SAFE]: The skill implements a structured, user-led writing workflow without any automated scripts or tool invocations. The instructions explicitly state that raw material files are read-only and that the AI should only append to files with user consent. Metadata analysis shows no deceptive patterns, and the disable-model-invocation: true flag in the frontmatter further restricts the agent's ability to trigger external tools autonomously. The process focuses on human-in-the-loop content generation, mitigating risks associated with autonomous agent behavior.

writing-shape