Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

ai-agent-book

ai-agent-book: An open-source Chinese book that explains AI Agents from design principles to engineering practice through the formula Agent = LLM + context + tools.

What problem it solves

Readers need a structured way to understand AI Agent design principles and apply them through engineering practice. The book addresses this by organizing the material into 10 chapters and pairing the text with 109 companion experiments covering topics such as context engineering, memory, tools, evaluation, post-training, continuous evolution, and multi-agent collaboration.

Key features

  • Ten-chapter progressive curriculum — The book is organized around one formula, Agent = LLM + context + tools, and progresses in ten chapters from agent basics and context engineering through tools, coding agents, interaction, evaluation, post-training, continuous evolution and multi-agent collaboration.

    Source: README, book, book/chapter10.md

  • Companion experiments per chapter — Each chapter ships a set of hands-on experiments with implementation and reproduction notes, mixing locally runnable projects with external reproduction tracks. Installations are scoped per chapter so readers only pull in the dependencies an experiment actually needs.

    Source: README, chapter1

  • Free PDF, EPUB and online reading — Full-text builds are published as PDF and EPUB for offline reading in fifteen languages, and an online version offers language switching, chapter collapsing, highlighting and notes. Chinese is the source of truth, with other languages contributed by the community.

    Source: README

  • Reproducible chapter-scoped dependency installs — Readers can sync a locked, reproducible environment for a given chapter with uv, or resolve dependencies from pyproject.toml with pip. Heavy stacks such as local serving and fine-tuning remain opt-in extras so a chapter install does not fail on CPU-only machines.

    Source: README, uv.lock

  • Community contribution and translation workflow — The project accepts pull requests for errata, code fixes, new practice projects, better SVG figures and additional translations. Because Chinese is the canonical text, content edits are made there and translations are synced after merge.

    Source: README, docs

Quick start

Copied verbatim from the project's README.

# 推荐:使用提交到仓库的 uv.lock,获得可复现的章节环境
uv sync --locked --extra ch1

# 未安装 uv 时:使用 pip 从 pyproject.toml 重新解析
python -m pip install -e ".[ch1]"
uv run python chapter1/context/main.py
# 使用 pip 安装时也可直接运行:python chapter1/context/main.py

Who it may suit

Inference: this may suit developers who can read and modify moderately complex Python and have used LLM products and AI-assisted coding tools. It may also suit teams building coding agents who want evaluation sets, memory and multi-agent patterns to implement from scratch.

Topics to discuss

  • Which chapter experiment would you most want to see rewritten as a runnable, CPU-friendly tutorial?

  • How do you decide when to reach for SFT versus RL when tool calling does not generalize?

  • What should a minimal evaluation set for a personal coding agent contain, and how should its tasks be scored?


Written by ReadmeX AI from the project's README, source and public GitHub data; it may contain mistakes.

Read the original README https://github.com/bojieli/ai-agent-book#readme

ReadmeX opened this community for a trending project; it is not run by the project. Maintainers can claim it from the community page.

github.comhttps://github.com/bojieli/ai-agent-book

Comments

…