noui-generalize

Generalize recorded NoUI workflows into agent-invokable FastMCP server tools.

25|8|Updated Apr 6, 2026
One-click install
npx skills add https://github.com/adoptai/noui --skill noui-generalize
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: noui-generalize
Source: https://github.com/adoptai/noui/tree/main/skills/noui-generalize
Command: npx skills add https://github.com/adoptai/noui --skill noui-generalize

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Recorded MCP workflows often produce tool signatures and execution behavior that are brittle, confusing to agents, or incompatible with the target site’s anti-bot protections, making the workflow hard to reuse reliably.

Core Features & Use Cases

  • Execution strategy hardening: Ensures generated servers reliably execute via Tabby’s browser-side execute/fetch flow, handling HITL-login edge cases and anti-bot failures.
  • Interface cleanup for agent usability: Renames opaque internal tool parameters into natural-language arguments so Claude Code can call tools without knowing implementation details.
  • Bot-detection and auth-failure mitigation: Guides targeted fixes for credential formatting issues, ACTIVE profile promotion, rate-limiting scenarios, and browser-auth execution modes.

Quick Start

Use the noui-generalize skill to generalize your recorded workflow into a Claude Code–usable tool interface while preserving the working browser execution mechanics.

Frequently Asked Questions about noui-generalize

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I turn recorded MCP workflows into usable tools for Claude Code?

To turn recorded MCP workflows into usable tools, you must rewrite opaque tool interfaces into natural-language parameters and harden the browser-side execution mechanics, ensuring the generated server reliably handles agent runtime invocations.

Why do generated FastMCP servers fail during agent execution?

Generated FastMCP servers fail during agent execution because recorded workflows often produce brittle tool signatures and encounter anti-bot protections or auth failures that block reliable browser-side execution.

How do I fix anti-bot detection and auth failures in MCP workflows?

Fixing anti-bot detection and auth failures in MCP workflows requires diagnosing browser execution strategy, correcting credential formatting, promoting ACTIVE profiles, and applying rate-limiting mitigations for browser-auth modes.

How do I make MCP tool parameters readable for agent runtimes?

Making MCP tool parameters readable for agent runtimes involves generalizing the recorded workflow by renaming opaque internal parameters into natural-language arguments so agents can invoke tools without knowing implementation details.

Does noui-generalize preserve browser execute-fetch mechanics?

Yes, noui-generalize preserves browser execute-fetch mechanics by hardcoding recorded infrastructure routing values and ensuring Tabby's browser-side execution flow remains intact while updating tool interfaces for agent usability.

What is the best way to update tools.json and server registration after generalizing a workflow?

The best way to update tools.json and server registration after generalizing a workflow is to consistently update both the server registration and documentation alongside the rewritten tool interfaces to maintain execution reliability.