Python Automation Cookbook 100+ new and updated recipes for scalable workflows, MCP integrations, and AI-powered automation (Jaime Buelta) (z-library.sk, 1lib.sk, z-lib.sk)

【One-Line Pitch】 A recipe-driven guide that takes Python users from everyday scripting chores—files, scraping, APIs, scheduling—into modern AI-assisted automation with model calls, MCP integration, and intelligent agents. Best for developers, DevOps engineers, and sysadmins who already know basic Python and want practical, copy-adaptable solutions. 【Book Arc】 - **Opening (~0%–10%)**: Sets up the automation mindset and toolchain—virtual environments, dependency pinning, string formatting, parsing, and using GenAI as a coding assistant—so later recipes run reproducibly. - **Early (~10%–30%)**: Covers command-line interfaces, scheduling with cron-style periodic tasks, email/notification delivery, and the basics of web data: RSS/Atom feeds and HTML form structures. - **Middle (~30%–50%)**: Moves into web scraping and crawling with concurrency, then local file search and reading across formats (text, CSV, PDF, DOCX, OCR), establishing the data-gathering layer. - **Late (~50%–75%)**: Focuses on turning raw data into outputs—report generation with templating (Jinja-style) and document creation (Word/PDF)—plus the testing and debugging discipline that keeps automation reliable. - **Ending (~75%–100%)**: Expands into AI-powered automation: calling AI models from scripts, implementing the Model Context Protocol (MCP) for system integration, and designing intelligent agents for business workflows. (Excerpts do not cover the detailed recipes of these final chapters.) 【Key Takeaways】 - **GenAI is treated as a coding collaborator, not an oracle** (Opening): prompts generate starter code and selectors, but the book stresses verifying results—especially type handling in parsing. (Early) - **Reproducibility starts with environments and pinned dependencies** (Opening): virtual environments plus requirements files are the baseline for every recipe. (Early) - **Scheduling is deceptively simple and operationally risky** (Early): cron-style jobs can't express sunset-relative or holiday-aware timing, and overlapping long tasks can cause race conditions or host overload; lock files and generous time budgets are the practical safeguards. (Early) - **Concurrency is the main lever for scraping speed** (Middle): worker-count experiments show crawl time dropping as parallelism rises, with the trade-off of added complexity and resource pressure. (Middle) - **Data gathering spans messy real-world formats** (Middle): recipes cover RSS/Atom feeds, HTML forms, local text/CSV, PDFs, DOCX, and OCR, so automation isn't limited to clean APIs. (Middle) - **Reporting is a first-class automation output** (Late): templating with loops/filters and programmatic document styling turn processed data into shareable artifacts. (Late) - **The new edition's differentiator is AI-native automation** (Ending): model calls, MCP for system-level integration, and agents that make decisions move scripts toward adaptive systems. (Ending) - **Testing and debugging are woven through, not bolted on** (Late): the book treats reliability as part of the recipe, not an afterthought. (Late) 【Reading Tips】 - **Skim the Opening if you're already comfortable with venvs and formatting**; deep-read the scheduling and concurrency sections, where the operational pitfalls are non-obvious. - **Treat recipes as templates, not scripts to run verbatim**—adapt paths, credentials, and selectors to your environment. - **Pause on the AI/MCP/agent chapters** even if they're unfamiliar; they're the reason to pick up this edition over earlier ones. - **Keep the requirements files handy** and reproduce each recipe in an isolated environment to avoid dependency drift. - **Note the "See also" cross-references**—they chain related recipes (e.g., parsing → regex → third-party tools) into a learning path. 【Coverage Limits】 This guide is synthesized from stratified excerpts covering roughly the first half of the book plus the blurb; the AI, MCP, and agent chapters are described at a high level only, and specific recipes, figures, or chapter titles from the later sections are not detailed here.