apify-brand-reputation-monitoring
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data (reviews, comments, and mentions) from external social media and review platforms via Apify Actors. This content is inherently untrusted and could contain malicious instructions intended to influence the agent's behavior during the summarization phase.
- Ingestion points: Data is fetched from the Apify API and processed in
reference/scripts/run_actor.jsbefore being displayed to the agent or saved to a file. - Boundary markers: The instructions recommend redacting personal fields and validating samples before labeling trends, which serves as a manual oversight boundary.
- Capability inventory: The script performs file writes (
node:fs) and network requests to Apify's API (fetch). The agent is tasked with summarizing the retrieved data. - Sanitization: The script implements robust sanitization for CSV exports, escaping characters (
=,+,@,-) that could trigger formula execution in spreadsheet applications (CSV injection protection). It also redacts theAPIFY_TOKENfrom error logs.
Audit Metadata