webmcp

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides developer guidance for integrating the WebMCP (Web Model Context Protocol) into browser-based applications. All technical instructions follow standard web development patterns and refer to legitimate upcoming browser APIs.
  • [SAFE]: The scripts/find-webmcp-targets.mjs script performs local filesystem scanning to identify relevant frontend files for integration. It uses standard Node.js fs and path modules with an appropriate ignore list (node_modules, .git, etc.) and does not perform any network operations.
  • [SAFE]: External URLs referenced in references/compatibility.md and references/webmcp-reference.md target official and well-known repositories, specifically GoogleChromeLabs/webmcp-tools and the W3C/WebML Community Group's webmachinelearning/webmcp. These are trusted sources for web standards and experimental browser features.
  • [SAFE]: The provided code templates (e.g., assets/model-context-registry.template.ts) use standard defensive programming techniques such as feature detection for experimental APIs (document.modelContext || navigator.modelContext) and proper cleanup using AbortController to manage tool lifecycles.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 07:58 PM
Security Audit — agent-trust-hub — webmcp