bilibili-danmaku-scraper

Warn

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill mandates downloading a dependency from an external, non-standard domain (browseract.com).
  • [REMOTE_CODE_EXECUTION]: The instructions command the agent to perform automatic dependency installation without user confirmation or approval, bypassing standard safety checkpoints.
  • [PROMPT_INJECTION]: The skill ingests untrusted user content (Bilibili danmaku) without sanitization or boundary markers, creating a surface for indirect prompt injection.
  • Ingestion points: Danmaku text from Bilibili video pages (SKILL.md).
  • Boundary markers: Absent; the skill does not instruct the agent to use delimiters or ignore instructions within the scraped data.
  • Capability inventory: Uses the browser-act tool for web interaction and data collection (SKILL.md).
  • Sanitization: Absent; no mention of filtering or escaping external content.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 20, 2026, 04:49 AM
Security Audit — agent-trust-hub — bilibili-danmaku-scraper