The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes shell command snippets for technical SEO auditing and WordPress management. This includes the use of curl to fetch webpage content, piping the output to python3 for JSON-LD extraction, and utilizing wp-cli for bulk database updates.
[INDIRECT_PROMPT_INJECTION]: The diagnostic workflows involve fetching and processing content from external, user-supplied URLs, creating a surface where untrusted data enters the agent's context. 1. Ingestion points: The curl command snippets in SKILL.md are designed to fetch HTML content from remote URLs. 2. Boundary markers: No delimiters or explicit instructions to ignore embedded prompts are provided in the diagnostic scripts. 3. Capability inventory: The skill utilizes network requests, shell access, and Python scripts for data extraction and manipulation. 4. Sanitization: The extraction logic uses regular expressions and standard JSON parsing, which prevents direct execution of the fetched content, but the agent's interpretation of the extracted data remains a potential vector.
[EXTERNAL_DOWNLOADS]: The provided diagnostic scripts facilitate the download of HTML data from remote servers (using placeholders like example.com) for the purpose of validating structured data.

schema-markup