monitoring-instagram-brand-mentions

Warn

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill definition includes an input parameter customMapFunction, described as a JavaScript function string used to transform output objects. This allows for the execution of dynamic logic within the data processing pipeline.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from Instagram, including post captions, comments, and user handles. This content is used for sentiment analysis and classification, creating a surface for indirect instructions to influence the agent's output.
  • Ingestion points: Data fetched from Instagram posts (captions, comments, author handles) via the Apify actor.
  • Boundary markers: None identified in the instructions for separating untrusted content from agent logic.
  • Capability inventory: Subprocess execution via node scripts/run_actor.js and network operations via curl.
  • Sanitization: No explicit sanitization, filtering, or escaping of the scraped Instagram content is mentioned before it is processed or reported.
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to execute shell commands using node for local script execution and curl for interacting with the Apify REST API.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 2, 2026, 07:32 PM
Security Audit — agent-trust-hub — monitoring-instagram-brand-mentions