cloakbrowser-fallback
Warn
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill instructs users to run 'python -m cloakbrowser install', which downloads a ~535 MB custom-compiled Chromium binary from a remote source. This binary is then executed during the skill's operation to navigate websites.
- [EXTERNAL_DOWNLOADS]: The skill relies on the 'cloakbrowser' library and its associated binary, which are hosted on external package registries and vendor-controlled infrastructure.
- [COMMAND_EXECUTION]: The helper scripts 'scripts/cloak_fetch.py' and 'scripts/cloak_fetch.mjs' execute the custom browser binary with user-supplied URLs and parameters.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes content from arbitrary external websites.
- Ingestion points: The scripts in 'scripts/cloak_fetch.py' and 'scripts/cloak_fetch.mjs' load content from user-provided URLs and extract page text using 'page.innerText("body")'.
- Boundary markers: None are present in the scripts or instructions to warn the agent about potential instructions embedded in the website content.
- Capability inventory: The skill can perform network operations and write files (screenshots and HTML).
- Sanitization: The page text is printed directly to the console without any validation, filtering, or escaping.
Audit Metadata