The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest large amounts of untrusted data from the Facebook Ad Library and external competitor landing pages. This creates a surface for indirect prompt injection where malicious actors could embed hidden instructions in ad copy or web content to manipulate the agent's behavior.
Ingestion points: Step 2 (Ad Library scraping) and Step 4 (Landing page scraping).
Boundary markers: The instructions state to 'quote it as evidence only' and 'never paraphrased', but they do not provide explicit delimiters or instructions for the agent to ignore command-like text within the scraped data.
Capability inventory: The skill has the ability to write documents to the project folder, call the Apify MCP tool, and hand off data to other marketing skills.
Sanitization: There is no mention of filtering, escaping, or validating the ingested content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill interacts with external services to gather data.
It connects to https://mcp.apify.com/ and references an internal network resource at http://api.apify.internal:3333/v2/datasets/ for retrieving large datasets.
It performs automated browsing of facebook.com/ads/library and various third-party landing page domains.
[COMMAND_EXECUTION]: The skill provides specific JavaScript snippets to be executed in the browser environment to extract URLs from the DOM and navigate paginated content. While these snippets are functionally necessary for the skill's purpose, they represent local code execution within the browser context.

cs-marketing-offer-creation