brandonfuchs.com

What It Is

AI Tooling Field Guide is a personal reference site built by Brandon Fuchs. Its goal is to untangle concepts in the AI tooling ecosystem that are often conflated: models, APIs, protocols, tools, skills, hooks, agent runtimes, governance, evaluation, and cost. It is not an AI SaaS product that generates content directly, but rather a learning map and experimental handbook for developers.

Core Capabilities and Content

The site’s core value lies in building an engineering-oriented mental model. The text repeatedly emphasizes that a model is simply a “text in, text out” program; the real complexity sits around access paths, context, tool calls, protocol adaptation, and security boundaries. It covers ways to access models, including subscription products, Provider APIs, hosted model platforms, routers, local runtimes, and model files. It also explains topics such as MCP, OpenAPI, function calling, LSP, Agent loops, RAG, Evals, Fine-tuning, and Cost & Tokens. The Labs section provides a hands-on path from Lab 00 to Lab 19, requiring the use of a terminal and Python 3. It is suitable for learning CLI workflows, JSON interfaces, MCP servers, memory graphs, logging, and evaluation through practice.

Pricing and Access

The crawled text does not show any pricing, subscription, or paid course information for the site. Overall, it appears to be a public personal reference resource. Its Cost & Tokens page discusses token-based billing logic for third-party models and gives pricing examples from OpenAI, Anthropic, Google, and others, but clearly notes that these are only rough references and that users should check official pricing pages. Access from mainland China is not reflected in the text, and there is no information about payment methods.

Pros and Cons

Its strengths are that the content is restrained and engineering-focused. It does not mythologize agents or tool protocols, and it helps readers determine which layer a new tool belongs to: model access, tools, protocols, runtime, or governance. It also explains token costs, context windows, caching, and the differences between APIs and local deployment clearly. The downside is that it is not a mature product: there is no information about APIs, enterprise support, SLAs, or privacy terms. The author also states that it is a living document, so some pages may be thin, incorrect, or outdated. Chinese-language support is not mentioned.

Who It’s For

It is suitable for software engineers, AI application developers, technical leads, and anyone who wants a systematic introduction to the AI tooling stack. If your goal is to find a ready-to-use chatbot, writing tool, or commercially available platform for the China market, this is not a direct substitute. Alternatives include OpenAI, Anthropic, Gemini, LangChain, and LlamaIndex; in China, you can refer to platform documentation from Tongyi Qianwen, Zhipu AI, Baidu Qianfan, Volcano Ark, and others.