Zukko Blog
Guides to Claude Code, Cursor, Codex and the Claude and GPT APIs: setup, token savings and troubleshooting.
AI for Word documents: create and edit DOCX through an API
4 min · · автоматизация
A model cannot save a file into Word, but it can return structure and text. A practical python-docx plus API pipeline for creating and editing documents.
LangChain: what it is, what it consists of and when you really need it
5 min · · development
What LangChain is and which packages it consists of, which tasks it helps with and when the plain OpenAI SDK is enough. We compare code and spending.
OpenAI SDK: what it is and why it works with more than GPT models
5 min · · development
The OpenAI SDK is an ordinary wrapper around an HTTP API. That is why you can point it at another address and use Claude, Gemini or DeepSeek without changing your code. Here is how, and where the limits are.
RAG explained in plain words and when you really need it
5 min · · models
RAG adds found fragments of your documents to the request to the model. We explain the design without jargon and calculate when long context is simpler.
Claude Code: what this agent is and how it differs from a Claude chat
6 min · · агенты
Claude Code works not in a chat window but in your project folder: it reads files, edits them and runs commands. What it can do and what it costs in tokens.
OpenAI API key example: using a key in curl, Python and Node.js
5 min · · API keys
Ready examples of passing an API key in a header, environment variables and an SDK, the typical format mistakes, and a key check with a /v1/balance request.
Vibe coding in Russia: which tools work and what they cost
6 min · · агенты
Which vibe coding tools you can connect to models through a gateway, how to set them up and what an hour of agent work costs.
OpenAI Python SDK: timeouts, retries, errors and raw responses
5 min · · development
The openai client for Python does more than create(). We cover timeouts, retries, exceptions, raw responses and streaming, using a gateway as the example.
AI document translation for Word files that keeps formatting
4 min · · автоматизация
Sending a whole DOCX to a model as plain text loses the formatting. We translate paragraph by paragraph, put the text back in place and estimate the cost.
Free AI for programming: what is actually available
5 min · · development
Free AI coding tools exist, but each one has a price in limits, hardware or time. We go through the options and calculate what a paid minimum costs.
An MCP server for 1C: giving an agent your configuration metadata
6 min · · агенты
Without MCP the model invents attributes your configuration does not have. We show what an MCP server for 1C provides, how to install mcp-1c and connect an agent through a gateway.
Cursor AI: what this editor is and what it can do in 2026
6 min · · IDE
Cursor is a code editor with an AI agent, chat and autocomplete. We explain what it consists of, how it differs from plain VS Code and when to connect your own key.
Translating PDFs from English to Russian with AI: text, tables, formulas
5 min · · автоматизация
A working PDF translation pipeline: extract the text, split it, translate with a glossary and assemble the result. What 100 pages cost.
An AI model to write website code: from mockup to a working page
5 min · · development
An agent can build a one-page site in one session if you brief it well and check the result in a browser. The process and the price, step by step.
LangChain vs LangGraph: when you need chains and when you need a graph
5 min · · development
LangGraph does not compete with LangChain, it sits underneath. We cover when a chain or create_agent is enough, when you need a stateful graph, with code on one model.
Function calling: how an agent works with external systems
5 min · · агенты
The model calls nothing itself: it only asks your program to run a function. The whole loop step by step, with Python code and a price estimate.
LlamaIndex vs LangChain: what to pick for RAG and agents in 2026
5 min · · development
Both libraries do RAG and agents but grow from different ideas. We compare the approaches, show minimal code with one model and price a request.
OpenAI API key: what it is and how it differs from ChatGPT
5 min · · API keys
An API key is a secret string that tells the service who is calling and whose balance to charge for tokens. Here is how it differs from a ChatGPT account.
LLM integration in Laravel: HTTP client, service, streaming and queues
4 min · · development
Connecting a model to Laravel without third-party packages: the Http facade, a service class, a streaming response, background jobs with retries and Http::fake tests. With 401–503 errors explained.
Caching LLM responses in Redis: stop paying for repeated requests
6 min · · development
Zukko already serves an identical repeated request from cache for free. Here is when that is not enough and you need your own Redis: input normalization, long TTL, prompt versions, redis-py code.
Writing technical specs with AI: process, prompt, template
6 min · · development
A good spec from a model is the result of a good interview. Prompts for each step, API code, common mistakes and a checklist before handing it to a developer.
Summarizing Long Documents: Map-Reduce or a 1M Context Window
6 min · · development
Map-reduce is for documents that do not fit the context window or need parallelism, not for saving money. Both approaches with code and cost math.
Auto-Documenting Code with AI: Docstrings, README and CI
4 min · · development
A Python script finds functions without docstrings, asks a model to describe them, and inserts the text into the code safely. Plus pre-commit and CI setup.
Generating Commit Messages from a Diff with AI
5 min · · development
A prepare-commit-msg hook and a Python script that draft a Conventional Commits message from the staged diff. You edit it in your editor as usual.