story-long-scan

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on local Node.js scripts to automate data collection from novel platforms. These scripts execute a browser automation CLI using subprocess calls. The implementation includes logic to safely resolve executable paths on Windows to avoid shell-related issues.\n- [DYNAMIC_EXECUTION]: The scraping scripts utilize eval within the browser environment to extract data from the DOM. This dynamic execution is scoped to the web pages being scraped and is a standard mechanism for browser-based automation.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external novel platforms, creating an attack surface for indirect prompt injection.\n
  • Ingestion points: Public novel ranking data scraped from various platforms via the provided scripts.\n
  • Boundary markers: Ingested data is structured within Markdown reports using headers and lists.\n
  • Capability inventory: Subprocess execution for scraping tools and local file writing for reporting.\n
  • Sanitization: The scripts include logic to normalize and truncate description text before it is included in reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 03:18 PM
Security Audit — agent-trust-hub — story-long-scan