The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze content from external websites via the /seo ecommerce <url> command. This processing of untrusted third-party data represents an indirect prompt injection surface, as the agent could potentially follow malicious instructions hidden within the target page's HTML, metadata, or schema markup.
Ingestion points: SKILL.md describes a workflow using render_page.py and parse_html.py to fetch and parse external URLs.
Boundary markers: The documentation does not specify the use of delimiters or clear separation between ingested data and agent instructions.
Capability inventory: The skill possesses the ability to execute shell commands (via claude-seo run) and interact with external APIs.
Sanitization: There is no evidence of filtering or sanitization being applied to the data retrieved from external sources before it is processed by the agent.- [COMMAND_EXECUTION]: The skill orchestrates its functionality by executing several local Python scripts through the claude-seo command-line utility. This includes scripts for page rendering (render_page.py), SEO parsing (parse_html.py), cost verification (dataforseo_costs.py), and API communication (dataforseo_merchant.py).

seo-ecommerce