bun-publishing

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes user-controlled metadata from files like package.json, which serves as a potential vector for indirect prompt injection.
  • Ingestion points: package.json (name, version, description) and project source files.
  • Boundary markers: The skill does not specify delimiters to isolate user-controlled data from instructions.
  • Capability inventory: The skill uses Bash to execute npm and bun commands, and Write for file operations.
  • Sanitization: No sanitization of user-provided metadata is described before its use in shell commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 02:26 AM
Security Audit — agent-trust-hub — bun-publishing