pinterest-board-scraper

Warn

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references an external URL (https://www.browseract.com/...) for the installation of the BrowserAct CLI. The instructions specifically command the agent to install this dependency "automatically without asking the user to confirm" and to "not request separate approval," which explicitly bypasses user oversight and consent.
  • [COMMAND_EXECUTION]: The skill involves the execution of the browser-act CLI and instructs the agent to perform system-level installations of missing dependencies, which requires executing shell commands.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection by processing untrusted data from Pinterest board metadata and user-generated content.
  • Ingestion points: Board titles, captions, descriptions, hashtags, and comments fetched from Pinterest (SKILL.md).
  • Boundary markers: None. The skill does not provide delimiters or instructions to treat the scraped data as untrusted or distinct from the system prompt.
  • Capability inventory: The skill uses the BrowserAct CLI for web interaction and possesses the capability to download and install external software.
  • Sanitization: There is no evidence of data validation, escaping, or sanitization of the ingested content.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 20, 2026, 04:47 AM
Security Audit — agent-trust-hub — pinterest-board-scraper