WebMCP agent skill — make any web app agent-ready, verifiably: inventory, integrate, verify in a real browser, heal. Claude Code, Codex, Cursor, opencode, Copilot.
Generate safe, typed, human-reviewed WebMCP tools from the API contracts you already have, instead of hand-writing registerTool() calls for every action.
A WebMCP-native architecture studio where an engineer and their agent co-design a system, run it against traffic and failures, and turn the result into an explainable architecture—not a static diagram.
🧭 Agent-native career transition platform for Vietnam's workforce built on WebMCP. Exposes 12 browser-level tools for in-app AI agents (ChatGPT/Chrome), pairing verified Vietnam labor databases with human-in-the-loop plan approvals (requestUserInteraction). ✨
Browser-local data sonification studio. Turn CSV columns into deterministic, editable music with inspectable provenance.
WebMCP-native incident review studio — agents investigate, humans authorize.
A live 3D command center where an analyst and an AI agent investigate sensitive factory data together — built for the OpenAI WebMCP Challenge.
To associate your repository with the webmcp-agent-skill topic, visit your repo's landing page and select "manage topics."