The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions direct the agent to execute a local shell command using Node.js (node scripts/run_actor.js) to interact with the Apify platform. While this is the intended functionality for the user, it grants the agent the ability to execute script files within the skill's environment.
[EXTERNAL_DOWNLOADS]: The skill makes network requests to the Apify API (api.apify.com) and utilizes remote actors (apidojo/twitter-user-scraper, apidojo/tweet-scraper) to fetch data. These resources are associated with the skill's author and represent official integrations with a well-known service.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes untrusted data from Twitter (tweets, user bios).
Ingestion points: Data enters the context via the tweet-scraper and twitter-user-scraper outputs mentioned in the workflow steps.
Boundary markers: The instructions do not define explicit delimiters or instructions for the agent to ignore embedded commands within the scraped social media content.
Capability inventory: The skill has capabilities for shell execution (node scripts/run_actor.js), network operations (curl to Apify API), and file writing (via the --output flag in the script).
Sanitization: There is no mention of sanitizing or escaping the scraped content before it is processed by the scoring logic or formatted for output.
[DYNAMIC_EXECUTION]: The skill documentation includes an optional customMapFunction parameter, which is a string representing a JavaScript function. This function is intended to be passed to the remote Apify actor for runtime data transformation. While execution occurs on the remote scraping platform rather than the local agent environment, it constitutes a dynamic code generation pattern.

building-twitter-industry-watchlist