local-explorer

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • Development Command Execution: The skill includes standard build and code generation commands (e.g., pnpm generate:api, pnpm build). These are used to synchronize the local development environment with API specifications and are expected within a developer-oriented workflow.
  • Trusted External Reference: The documentation references Cloudflare's official Kumo repository on GitHub for UI component guidelines. This is a reference to a trusted organizational resource intended to ensure consistency with existing internal tools.
  • Local Filesystem Access: The instructions describe using a loopback service (MINIFLARE_LOOPBACK) to access the host filesystem from within the worker runtime. This is a design requirement for the Local Explorer to manage local state files (like Durable Object state or KV namespaces) during development.
  • Data Ingestion Surface: The skill details how the Local Explorer aggregates data from user resources (KV, R2, D1). While this creates a surface where the agent might process user-controlled data, the skill includes instructions for validation using Zod schemas and standardized response envelopes, which are standard measures for maintaining data integrity.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 11:02 AM
Security Audit — agent-trust-hub — local-explorer