The Agent Skills Directory

[DATA_EXPOSURE]: The skill reads from local files such as docs/seo/brand-voice.md and docs/seo/topical-clusters.md to gather context for generating research prompts.
[COMMAND_EXECUTION]: The skill writes generated content to the local file system at docs/seo/<slug>/prompt.md. This file-writing behavior is consistent with the skill's stated purpose of organizing SEO research materials.
[PROMPT_INJECTION]: The skill includes instructions to generate a prompt containing 'IMPORTANT' anti-hallucination directives. These are part of the target output for a separate AI session (Claude.ai) and do not represent an attempt to bypass the current agent's safety protocols.
[DATA_EXPOSURE]: The skill processes user-provided topics and local project files to generate content. It does not perform network operations or access sensitive system credentials.

seo-research