lovable-to-nextjs

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted user-provided source code to perform diagnostics and validation, which can serve as a vector for indirect prompt injection if the code contains adversarial instructions.
  • Ingestion points: The scripts scripts/detect-lovable-stack.ps1, scripts/quality-diagnostic.ps1, and scripts/validate-migration.ps1 read project files including package.json, .tsx, .ts, .css, and index.html into the agent's context.
  • Boundary markers: The instructions do not define clear boundary markers or protective delimiters to isolate ingested code content from the agent's core instructions during the multi-phase pipeline.
  • Capability inventory: The skill possesses powerful capabilities including Bash (shell execution), Write, and Edit, allowing it to modify the project structure and files based on its analysis.
  • Sanitization: There are no documented procedures for sanitizing or validating the content of ingested source files before processing them, relying on the LLM's internal safety filters.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 01:49 PM
Security Audit — agent-trust-hub — lovable-to-nextjs