The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses curl and grep to audit website metadata by fetching HTML content from various domains associated with the author, such as magnus919.com and rdumesh.org.
[COMMAND_EXECUTION]: It utilizes the ghost-cli command-line tool to manage site settings, inject JSON-LD structured data, and update post metadata on Ghost CMS installations.
[EXTERNAL_DOWNLOADS]: The reference documentation describes installing the llms-txt Python package via pip to handle the generation of LLM-friendly site context files.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates an attack surface for indirect prompt injection as it ingests untrusted HTML content from external websites.
Ingestion points: Site pages and sitemaps are fetched using curl in SKILL.md to audit search visibility.
Boundary markers: None are present to delimit the external content from the agent's instructions, though the use of grep for specific tags limits the impact.
Capability inventory: The skill has the ability to execute shell commands (curl, ghost-cli), perform network operations, and modify site content via the Ghost CMS management tools.
Sanitization: There is no explicit sanitization or filtering of the fetched HTML content before it is inspected by the agent.

seo-content-optimization