apify-scrapers
Fail
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: CRITICALINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [METADATA_POISONING]: An automated malware scanner flagged
SKILL.mdwith aFileRepMalwaredetection. While the file content consists of markdown instructions and environment configuration guidance, the reputation-based flag suggests potentially malicious or unwanted patterns recognized by the scanner. - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of large amounts of untrusted content from the web and social media, creating a significant indirect prompt injection surface.
- Ingestion points: Multiple scripts fetch raw data from external sources, including
scripts/scrape_content_by_url.py(fetching tweets, YouTube data, and websites),scripts/scrape_multi_platform.py(TikTok, YouTube, and website content), and platform-specific scripts for Facebook, Instagram, LinkedIn, and Reddit. - Boundary markers: Absent. The scraped content (posts, comments, reviews, and page markdown) is parsed and returned to the agent context as structured JSON or plain text without clear delimiters or instructions to the LLM to treat the content as untrusted data.
- Capability inventory: The skill possesses significant capabilities, including network access (via Apify and Firecrawl APIs), local file system write access (saving results to
.tmp/), and the ability to execute the included Python scripts which perform these operations. - Sanitization: Absent. The scripts do not perform sanitization, filtering, or escaping of the retrieved web content to prevent control token injection or command instruction leakage into the agent's prompt context.
Recommendations
- CRITICAL: 1 file(s) identified as malware by FileRep - DO NOT USE
- AI detected serious security threats
Audit Metadata