The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the mcpc CLI tool and a provided Node.js script (run_actor.js). these operations are used to fetch actor metadata and execute scraping jobs through the Apify platform.
[EXTERNAL_DOWNLOADS]: Interacts with Apify's official API (api.apify.com) to start actor runs and retrieve scraped data. Apify is a well-known and established service for web automation and data extraction.
[INDIRECT_PROMPT_INJECTION]: As the skill is designed to ingest and summarize data from various external web platforms, it possesses an indirect prompt injection attack surface.
Ingestion points: Untrusted data from social media platforms and websites is processed in SKILL.md (Step 5) and run_actor.js.
Boundary markers: Absent; the skill does not define explicit structural delimiters (e.g., XML tags) to wrap external content during summarization.
Capability inventory: The skill can perform network operations, write results to the local filesystem (run_actor.js), and execute CLI tools.
Sanitization: Present; the run_actor.js script implements a csvCell function that escapes special characters (=, +, @, -) to prevent CSV/formula injection and truncates long text fields to limit context injection impact.
[SAFE]: The skill implements secure secret management by instructing the user to store the APIFY_TOKEN in a .env file, which is then loaded via Node.js native environment support rather than being hardcoded or passed in plain text.

apify-ultimate-scraper