next-dev-loop

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Local Command Execution: The skill utilizes shell commands to manage browser sessions and persist state during development. This includes setting environment variables and executing browser automation commands to verify application behavior.
  • External Package Installation: The instructions include commands to install or upgrade dependencies such as next and agent-browser. These tools are necessary for the skill's core functionality of verifying runtime changes in a Next.js environment.
  • Indirect Prompt Injection Surface: The skill is designed to ingest and analyze runtime data from the target application, including DOM content, console logs, and framework-level metadata. This ingestion is a functional requirement for verifying application behavior.
  • Ingestion points: Application output via agent-browser (DOM, console logs) and Next.js internal metadata via the /_next/mcp endpoint.
  • Boundary markers: No specific delimiters or "ignore instructions" warnings are applied to the ingested application data.
  • Capability inventory: The skill uses browser automation and local shell execution to perform its tasks.
  • Sanitization: No explicit sanitization or filtering of the application's runtime output was identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 10:47 AM
Security Audit — agent-trust-hub — next-dev-loop