The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several local Python scripts (e.g., scripts/fetch_page.py, scripts/dataforseo_merchant.py) to execute its workflows. These operations are restricted to the local environment and are used to process user-provided URLs and keywords.
[EXTERNAL_DOWNLOADS]: The skill interacts with the DataForSEO Merchant API to retrieve marketplace data from Google Shopping and Amazon. This is the primary function of the skill and is protected by a mandatory cost-check mechanism (scripts/dataforseo_costs.py) that requires user approval for high-cost or specific marketplace endpoints.
[DATA_EXFILTRATION]: No patterns of unauthorized data exfiltration were detected. The network operations are limited to well-defined SEO data providers and user-specified product pages.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external URLs via fetch_page.py and parse_html.py for SEO analysis. While this represents a potential surface for indirect prompt injection (e.g., malicious instructions hidden in a product page's HTML), the agent's tasks are scoped to extracting specific SEO elements like schema, headings, and metadata, which reduces the risk of malicious instruction execution.
Ingestion points: External HTML content is ingested through the fetch_page.py script (called in SKILL.md).
Boundary markers: None explicitly defined in the workflow description.
Capability inventory: Limited to file reading (HTML parsing) and API network calls; no dynamic code execution (eval/exec) is present.
Sanitization: The skill uses a dedicated parsing script (scripts/parse_html.py) to extract structured SEO elements, providing a layer of normalization for the ingested data.

seo-ecommerce