The Agent Skills Directory

[SAFE]: The skill performs its stated purpose of auditing ad-to-landing page message match without any detected security violations.
[COMMAND_EXECUTION]: The instructions mention the use of meta-ad-scraper and google-ad-scraper for inventorying ads. These are referenced as tools for the agent to execute to gather data on the user's domain.
[DATA_EXPOSURE]: The skill requires an Apify API token for scraping functionality. The instructions guide the user on its requirement rather than hardcoding any credentials.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes untrusted content from external landing pages.
Ingestion points: Landing page content is fetched via fetch_webpage and ad data is ingested via scrapers or CSV uploads.
Boundary markers: No specific delimiters or instructions to ignore embedded commands within the fetched content were identified in the instructions.
Capability inventory: The skill has the capability to write files to the local system (saving reports to clients/ directory) and perform network requests.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from external URLs before it is processed by the model.

ad-to-landing-page-auditor