bun

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for downloading the Bun runtime from its official website and using the oven-sh/setup-bun GitHub Action for CI environments.
  • [REMOTE_CODE_EXECUTION]: Documentation includes the standard and official installation method which involves piping a remote shell script from bun.com to the system shell.
  • [COMMAND_EXECUTION]: The agent is instructed to use the bun and bunx CLI tools for tasks such as running tests, building executables, and managing dependencies.
  • [PROMPT_INJECTION]: The skill identifies ingestion points in local files like package.json and .env, noting a capability inventory that includes shell execution and file writes; however, it specifies that Bun's shell API provides native sanitization via template interpolation to prevent injection.
  • [SAFE]: All external resources and commands described in the skill are consistent with the official Bun ecosystem and represent standard, secure development practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 07:58 PM
Security Audit — agent-trust-hub — bun