nextjs-dev-orchestration

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill acts as an orchestrator that ingests untrusted user requirements to generate instructions for sub-skills (nextjs-data-provider and nextjs-ui-dev). It lacks boundary markers or sanitization when interpolating this external data into the handoff templates, creating a surface for indirect prompt injection where maliciously crafted requirements could override sub-skill instructions. * Ingestion points: User-provided feature scope, data entities, and architecture requirements in SKILL.md (Steps 1, 4, and 6). * Boundary markers: Absent from the delegation templates used to hand off tasks to sub-skills. * Capability inventory: The skill coordinates file system modifications, API hook definitions, and business logic implementation through delegated skills. * Sanitization: No validation or escaping of user-provided text before interpolation into prompt templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 04:14 AM
Security Audit — agent-trust-hub — nextjs-dev-orchestration