Skills
Navigate
skills/unbrowse-ai/sites/wordunscrambler-me/Gen Agent Trust Hub

wordunscrambler-me

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill connects to a remote service at https://unbrowse.ai to facilitate communication with wordunscrambler.me.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted web content, which presents a surface for potential indirect prompt injection.
  • Ingestion points: Data is ingested via wordunscrambler_me__render_page and wordunscrambler_me__read_page tools defined in SKILL.md.
  • Boundary markers: No specific delimiters are used to wrap the retrieved web content in the agent's context.
  • Capability inventory: The skill allows for web page navigation, text extraction, and browser-based actions scoped to the wordunscrambler.me domain.
  • Sanitization: The instructions do not indicate that content retrieved from the external site is sanitized or filtered.
  • [DYNAMIC_EXECUTION]: The documentation describes a feature where user actions recorded via unbrowse.browse.act can be compiled into new agent tools by the remote platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 10:56 AM