bun-runtime
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious instructions, obfuscated code, or unauthorized data access patterns were identified across the skill and its reference files. The content serves as standard developer documentation for the Bun runtime.
- [COMMAND_EXECUTION]: The skill documents the capability to execute code via the Bun CLI (e.g.,
bun run,bun --eval). These are intended functionalities of the runtime environment and are presented in a neutral, informative context for development tasks. - [PROMPT_INJECTION]: The skill outlines usage patterns for a code execution environment, which constitutes an indirect prompt injection surface if used by an agent to process untrusted code.
- Ingestion points: The skill facilitates running code from local files (
bun run) and piping content through standard input (bun run -). - Boundary markers: There are no instructions for implementing delimiters or security boundaries when executing third-party or untrusted scripts.
- Capability inventory: The Bun runtime has comprehensive system access, including the ability to perform network requests (fetch), access the filesystem, and manage subprocesses.
- Sanitization: The instructions do not include methods for sanitizing or validating inputs before execution, relying on the runtime's default behavior.
Audit Metadata