The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx lighthouse@latest to execute performance and SEO audits. This involves downloading and running code from the official NPM registry, which is a standard method for this well-known tool from Google.
[COMMAND_EXECUTION]: The skill executes local Python scripts, such as sitemap_discovery.py and agent_ux_check.py, using a project-specific CLI claude-seo run. These scripts perform the specialized analysis required for technical SEO audits and are consistent with the skill's stated functionality.
[PROMPT_INJECTION]: The skill processes data from external websites, creating a surface for indirect prompt injection. 1. Ingestion points: website HTML, robots.txt, and accessibility trees fetched via render_page.py and agent_ux_check.py. 2. Boundary markers: No explicit delimiters or 'ignore' instructions are defined for the ingested content. 3. Capability inventory: Includes command execution via the claude-seo wrapper and file access for audit analysis. 4. Sanitization: No specific escaping or filtering of external content is described in the provided instructions. These risks are inherent to web analysis tasks and are typically mitigated by platform-level guardrails.

seo-technical