scraping-instagram-comments

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data from Instagram comments, which could potentially contain malicious instructions.\n
  • Ingestion points: The skill fetches comment text ('message'), user names, and full names from external Instagram posts (SKILL.md).\n
  • Boundary markers: There are no instructions provided to the agent regarding the use of delimiters or the need to ignore instructions embedded within the scraped content.\n
  • Capability inventory: The skill instructions include command line execution ('node scripts/run_actor.js') and network requests via 'curl' to external APIs.\n
  • Sanitization: No sanitization, escaping, or validation of the fetched comment content is mentioned.\n- [DYNAMIC_EXECUTION]: The 'customMapFunction' input parameter allows for the provision of a JavaScript function as a string to transform data objects at runtime.\n- [COMMAND_EXECUTION]: The documentation provides examples for running the scraper using a local script ('node scripts/run_actor.js'), which implies the execution of code from an external file.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 07:04 PM
Security Audit — agent-trust-hub — scraping-instagram-comments