The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by scraping external landing pages belonging to competitors. These pages could contain hidden instructions intended to manipulate the agent's behavior during the teardown and offer creation process.
Ingestion points: Step 4 ("Scrape the pages behind the winners") fetches content from external destination URLs identified in the Meta Ad Library.
Boundary markers: The skill instructions do not specify any delimiters or boundary markers to isolate the untrusted external content from the agent's internal instructions.
Capability inventory: The skill possesses the capability to read external web content and write durable documents to the project folder (Step 7).
Sanitization: There is no evidence of content sanitization or instruction to filter out potentially malicious text from the scraped pages.
[EXTERNAL_DOWNLOADS]: The skill references and downloads data from external services, including well-known platforms.
Evidence: Accesses public data from facebook.com/ads/library for competitor research.
Evidence: References the Apify platform via its MCP server at https://mcp.apify.com/ and accesses internal datasets through http://api.apify.internal:3333 for high-volume research.
[DYNAMIC_EXECUTION]: The skill provides specific JavaScript templates for the agent to execute within a browser environment to extract and decode URLs.
Evidence: Includes a JavaScript snippet in Step 2 to query the DOM for Facebook redirect links and decode the final destination URLs.

cs-marketing-offer-creation