The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes local Python scripts, scripts/render_page.py and scripts/parse_html.py, to facilitate the data acquisition and parsing phase. These scripts are executed via the command line with varying modes to handle SPA-aware rendering.
[EXTERNAL_DOWNLOADS]: The skill interacts with external services to gather search data. It uses the WebSearch tool to query Google and provides optional integration with the DataForSEO API (specifically the serp_organic_live_advanced and kw_data_google_ads_search_volume tools) for search metrics and live SERP data.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because its core function requires fetching and analyzing external content from user-provided URLs.
Ingestion points: Untrusted content enters the agent's context through scripts/render_page.py during Step 1 (Target Acquisition).
Boundary markers: The instructions do not specify the use of delimiters or specific framing to prevent the agent from obeying instructions that might be embedded within the target page's HTML or text.
Capability inventory: The agent has access to web search tools and the ability to interact with the DataForSEO API based on the findings from the untrusted data.
Sanitization: The skill documentation does not mention sanitization, filtering, or validation of the fetched content beyond basic structural parsing.

seo-sxo