The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill defines a customMapFunction input parameter described as a "JavaScript function to transform each output object." Accepting raw code strings for runtime execution is a security risk, as it may allow for arbitrary code execution within the processing environment if not strictly sandboxed.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process large amounts of untrusted data from Twitter (tweets, user bios, search results). This creates a vulnerability where malicious instructions embedded in tweets could influence the agent's behavior during the classification or scoring phases.
Ingestion points: Twitter search results and user metadata fetched via Apify actors (SKILL.md).
Boundary markers: None specified; the skill does not mention using delimiters or specific instructions to ignore content within the scraped data.
Capability inventory: The skill utilizes shell commands (node), network requests (curl), and file system operations (saving CSV/JSON).
Sanitization: No sanitization or filtering of the scraped content is mentioned before it is passed to the scoring and classification logic.
[COMMAND_EXECUTION]: The documentation provides multiple examples of shell command execution, including running local Node.js scripts (scripts/run_actor.js) and using curl to interact with the Apify REST API. These commands are used to manage data flow and actor execution.
[EXTERNAL_DOWNLOADS]: The skill fetches data from Apify's official API (api.apify.com) and utilizes specific Apify actors (apidojo/tweet-scraper, apidojo/twitter-user-scraper). These are recognized as well-known service endpoints and vendor-owned resources, representing the intended functionality of the skill.

monitoring-vc-investor-activity-on-twitter