scrape-structured-data

Warn

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEDYNAMIC_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install a global NPM package named chrome-agent if it is not already present on the system.
  • [CREDENTIALS_UNSAFE]: The skill documentation highlights a --copy-cookies flag used to read session cookies directly from the user's local Chrome profile to access authenticated web pages.
  • [DYNAMIC_EXECUTION]: The skill provides an eval command that allows the agent to execute arbitrary JavaScript code within the context of the browser instance it is controlling.
  • [COMMAND_EXECUTION]: The skill relies on executing various shell commands using the chrome-agent CLI to navigate web pages, extract data, and monitor network activity.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted websites, which presents an attack surface for indirect prompt injection.
  • Ingestion points: Data is extracted from arbitrary URLs via the extract, read, text, and network commands.
  • Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the scraped web content.
  • Capability inventory: The agent has the ability to execute shell commands, install software, and run JavaScript in a browser context.
  • Sanitization: While the tool outputs data in a structured JSON format, the textual content of the fields is directly derived from untrusted web sources without explicit sanitization steps described in the skill instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 2, 2026, 02:11 AM
Security Audit — agent-trust-hub — scrape-structured-data