bun-shell

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents the use of Bun's shell APIs (Bun.$, Bun.spawn, Bun.spawnSync) for executing system commands. The examples provided are standard development tasks, such as git operations, running build scripts, and managing files, which are consistent with the skill's stated purpose as a shell scripting utility.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a programming interface that can process external data through shell commands, creating a potential attack surface if used incorrectly.
  • Ingestion points: Demonstrated through variable interpolation in shell template literals in SKILL.md.
  • Boundary markers: Relies on Bun's runtime behavior which automatically escapes interpolated variables in Bun.$ templates to prevent shell injection.
  • Capability inventory: Detailed subprocess management including full stdio control, environment variable manipulation, and process signaling across all code examples.
  • Sanitization: The documentation explicitly teaches and recommends 'Safe interpolation (escaped)' as a core feature of the runtime.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 12:00 AM
Security Audit — agent-trust-hub — bun-shell