auditing-npm-dependencies

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes vulnerability advisories from the external npm registry. Maliciously crafted package metadata or advisory descriptions could theoretically contain prompt instructions designed to influence the agent's behavior during the audit process.
  • Ingestion points: The scripts/audit_npm.py script parses JSON output from the npm audit command.
  • Boundary markers: Absent. The skill does not implement specific delimiters or 'ignore' instructions for the data retrieved from the registry.
  • Capability inventory: The skill is authorized to use Bash for npm and python3 commands and can read project files.
  • Sanitization: Absent. The script maps external data directly to report fields without sanitizing for potential LLM instructions.
  • [COMMAND_EXECUTION]: The skill utilizes subprocess.run within scripts/audit_npm.py to execute the npm audit command. While this is the intended and necessary behavior for the skill's purpose, it represents an automated interaction with the system shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:41 AM
Security Audit — agent-trust-hub — auditing-npm-dependencies