setup

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides links to download the Vibe extension from the official Chrome Web Store and the author's GitHub repository (VibeTechnologies/VibeWebAgent).- [REMOTE_CODE_EXECUTION]: Instructs the user to run npx -y @vibebrowser/cli@latest, which downloads and executes the latest version of the vendor's CLI tool from the NPM registry to verify the connection.- [COMMAND_EXECUTION]: Uses shell commands to interact with the browser MCP server to list open tabs and confirm that the extension and relay are correctly configured.- [CREDENTIALS_UNSAFE]: References a 'remote UUID' used as a bearer credential for remote browser access. The skill provides proactive guidance on the safe handling of this secret, including warnings against sharing it or storing it in version-controlled files.- [PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by enabling tools that ingest and act upon untrusted browser content. * Ingestion points: Content is read from the browser via tools like snapshot and get_text. * Boundary markers: The instructions do not specify the use of delimiters or warnings for the agent when processing site content. * Capability inventory: The setup enables tools that can perform actions like click, fill, type, and eval within the browser context. * Sanitization: No sanitization or content filtering mechanisms are described in the setup process.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 07:05 PM
Security Audit — agent-trust-hub — setup