Book Library

How to be more ambitious: A tutorial for humans with models
How to be more ambitious: A tutorial for humans with models
6 pages
人人都是大厨：告别“抽盲盒”式下厨，讲透 22 条做菜的基本原理
人人都是大厨：告别“抽盲盒”式下厨，讲透 22 条做菜的基本原理
11 pages
You're Already a Meat Proxy
You're Already a Meat Proxy
7 pages
The Next Scaling Problem — Tetral
The Next Scaling Problem — Tetral
28 pages
Recommended Reading: catching up on Armin (thread)
Recommended Reading: catching up on Armin (thread)
1 pages
Recommended Reading (thread)
Recommended Reading (thread)
2 pages
我是如何构建一个 AI 原生量化系统的 · BMPI
我是如何构建一个 AI 原生量化系统的 · BMPI
15 pages
置身 Agent 时代：当软件开始不再需要软件工程师 · BMPI
置身 Agent 时代：当软件开始不再需要软件工程师 · BMPI
12 pages
Don’t Build Multi-Agents | Cognition
Don’t Build Multi-Agents | Cognition
9 pages
今天学习了一下amp团队是如何用agent的非常棒！特别适合刚入门agent的新
今天学习了一下amp团队是如何用agent的非常棒！特别适合刚入门agent的新
1 pages
LLM 为什么缓存 K 和 V，却通常不缓存 Q？
LLM 为什么缓存 K 和 V，却通常不缓存 Q？
4 pages
How to Run a 24/7 AI Company for $50/Month (GPT-6 Astra + DeepSeek V4.1 Flash - Full Setup Guide)
How to Run a 24/7 AI Company for $50/Month (GPT-6 Astra + DeepSeek V4.1 Flash - Full Setup Guide)
10 pages
The Future is Agentalization | Life Odyssey
The Future is Agentalization | Life Odyssey
25 pages
The Hitchhiker's Guide to Agentic AI: From Foundations to Systems
The Hitchhiker's Guide to Agentic AI: From Foundations to Systems
Haggai Roitman · 603 pages
LLM Powered Autonomous Agents | Lil'Log
LLM Powered Autonomous Agents | Lil'Log
Function calling | OpenAI API
Function calling | OpenAI API
Getting started with loops | Claude by Anthropic
Getting started with loops | Claude by Anthropic
Harness engineering: leveraging Codex in an agent-first world | OpenAI
Harness engineering: leveraging Codex in an agent-first world | OpenAI
How to Build an Agent Harness: A Practical Guide from Teams Who Actually Did It | Study Notes
How to Build an Agent Harness: A Practical Guide from Teams Who Actually Did It | Study Notes
Patterns for Building LLM-based Systems & Products
Patterns for Building LLM-based Systems & Products
A global workspace in language models | Anthropic
A global workspace in language models | Anthropic
Verbalizable Representations Form a Global Workspace in Language Models
Verbalizable Representations Form a Global Workspace in Language Models
How to Build the Loops That Just Replaced Entire Prompt Engineering
How to Build the Loops That Just Replaced Entire Prompt Engineering
Prompting guidance for GPT-5.6 Sol | OpenAI API
Prompting guidance for GPT-5.6 Sol | OpenAI API
harness engineering 101
harness engineering 101
How we built our knowledge base
How we built our knowledge base
From Loop Engineering to Graph Engineering?
From Loop Engineering to Graph Engineering?
17 prompts that make Hermes run while you sleep (copy-paste inside)
17 prompts that make Hermes run while you sleep (copy-paste inside)
Top 10 ML skills every AI engineer should know
Top 10 ML skills every AI engineer should know
Language model harnesses are compositional generalizers | Alex L. Zhang
Language model harnesses are compositional generalizers | Alex L. Zhang
Agent swarms and the new model economics · Cursor
Agent swarms and the new model economics · Cursor
Reliable unreliability | David Gasquez
Reliable unreliability | David Gasquez
Graph Engineering with Claude: 14-Step roadmap from 0 to graph architect
Graph Engineering with Claude: 14-Step roadmap from 0 to graph architect
12 Hermes Articles from 100+ of Hours Spent
12 Hermes Articles from 100+ of Hours Spent
Loop and Harness engineering: 7 files, 5 steps. Every config inside
Loop and Harness engineering: 7 files, 5 steps. Every config inside
State machines in 2 minutes
State machines in 2 minutes
The actor model in 2 minutes
The actor model in 2 minutes
Swarm Intelligence in Financial Market Analysis
Swarm Intelligence in Financial Market Analysis
Graph Engineering replaced RAG at Microsoft, Stanford and Anthropic. Here's how it works.
Graph Engineering replaced RAG at Microsoft, Stanford and Anthropic. Here's how it works.
Agentic Retrieval-Augmented Generation: A Survey on Agentic RAG
Agentic Retrieval-Augmented Generation: A Survey on Agentic RAG
Aditi Singh; Abul Ehtesham; Saket Kumar; Tala Talaei Khoei; Athanasios V. Vasilakos · 42 pages
One Token Is Enough: Fingerprinting and Verifying Large Language Models from Single-Token Output Distributions
One Token Is Enough: Fingerprinting and Verifying Large Language Models from Single-Token Output Distributions
Tomas Bruckner · 9 pages
Kimi K3 + Graph Engineering: 85% lower token costs and 18% better accuracy. Full guide to building t
Kimi K3 + Graph Engineering: 85% lower token costs and 18% better accuracy. Full guide to building t
On Making
On Making
Pensieve — Memory Management for AI Agents That Actually Forget - DEV Community
Pensieve — Memory Management for AI Agents That Actually Forget - DEV Community
You only need the frontier model for one single edit
You only need the frontier model for one single edit
Prompt Caching In Agents | EARENDIL
Prompt Caching In Agents | EARENDIL
advanced-context-engineering-for-coding-agents/wsff.md at main · humanlayer/advanced-context-engineering-for-coding-agents · GitHub
advanced-context-engineering-for-coding-agents/wsff.md at main · humanlayer/advanced-context-engineering-for-coding-agents · GitHub
Loop Engineering: The Karpathy Method - and the workflow that just made it 5x better
Loop Engineering: The Karpathy Method - and the workflow that just made it 5x better
Context Engineering: the Karpathy-Cherny method that replaced prompting
Context Engineering: the Karpathy-Cherny method that replaced prompting
Graph Engineering: build 1000+ agent loops in one window, from one prompt (full 5-step course)
Graph Engineering: build 1000+ agent loops in one window, from one prompt (full 5-step course)
Why Software Factories Fail
Why Software Factories Fail
Why Software Factories Fail: Turning the lights back on
Why Software Factories Fail: Turning the lights back on
How to master graph engineering (Full Course)
How to master graph engineering (Full Course)
Inference Engines for LLMs & Local AI Hardware (2026 Edition)
Inference Engines for LLMs & Local AI Hardware (2026 Edition)
Master Agent Architecture: Unifying Harness, Loop, and Graph Engineering
Master Agent Architecture: Unifying Harness, Loop, and Graph Engineering
22580: From GPT2 to Kimi3, Explained
22580: From GPT2 to Kimi3, Explained
个人 AI 基础设施构建指南
个人 AI 基础设施构建指南
How (and why) to build agent-first apps
How (and why) to build agent-first apps
We rewrote our agent to run entirely in a Durable Object with Pi, Agents SDK and Code Mode
We rewrote our agent to run entirely in a Durable Object with Pi, Agents SDK and Code Mode
What's gone wrong with AI & labor — a thought experiment
What's gone wrong with AI & labor — a thought experiment
The harness is all you need (mostly)
The harness is all you need (mostly)
Agent Harness 原语分类
Agent Harness 原语分类
How ChatGPT Optimizes its Agent Loop: Harness, API, and Inference
How ChatGPT Optimizes its Agent Loop: Harness, API, and Inference
Pragmatic Leverage in the Software Factory
Pragmatic Leverage in the Software Factory
5 Pieces of Math Every AI Actually Runs On
5 Pieces of Math Every AI Actually Runs On
Overview of Agent Sandboxes in Practice (Ephemeral Sandbox Volume1 Part 0)
Overview of Agent Sandboxes in Practice (Ephemeral Sandbox Volume1 Part 0)
We are not cleared for takeoff. Yet.
We are not cleared for takeoff. Yet.
FDE介绍和入门：一个被 AI 重新激活的老岗位
FDE介绍和入门：一个被 AI 重新激活的老岗位
Details That Make Interfaces Feel Better
Details That Make Interfaces Feel Better
Devtools must be open source - exe.dev blog
Devtools must be open source - exe.dev blog
LLMs reward expertise
LLMs reward expertise
Pi, Minimal and Performant | EARENDIL
Pi, Minimal and Performant | EARENDIL
Smaller, faster, safer: running Kimi and GLM at scale | The Cloudflare Blog
Smaller, faster, safer: running Kimi and GLM at scale | The Cloudflare Blog
Pydantic fixed my Agent's Memory
Pydantic fixed my Agent's Memory
Humans missed 1 in 3 threats approving AI agent commands across 40,000 plays | Scale X
Humans missed 1 in 3 threats approving AI agent commands across 40,000 plays | Scale X
AddyOsmani.com - My LLM coding workflow going into 2026
AddyOsmani.com - My LLM coding workflow going into 2026
Anatomy of a harness: building a coding agent that can run for hours
Anatomy of a harness: building a coding agent that can run for hours
一个零基础开发者的硬件 DIY 入门指南
一个零基础开发者的硬件 DIY 入门指南
"Code was never the hard part" is an insult to all programmers — Senko Rašić
"Code was never the hard part" is an insult to all programmers — Senko Rašić
The End of No Code - exe.dev blog
The End of No Code - exe.dev blog
Eval Engineering: build the gate that lets your agents merge without you (full 6-step course)
Eval Engineering: build the gate that lets your agents merge without you (full 6-step course)
AI for Physics 全景图
AI for Physics 全景图
Algorithms by Jeff Erickson - one of the best algorithm books out there
Algorithms by Jeff Erickson - one of the best algorithm books out there
Cloudflare Computer：如何将 AI Agent 沙箱成本降低 80%
Cloudflare Computer：如何将 AI Agent 沙箱成本降低 80%
Intro to Grok Bot
Intro to Grok Bot
DeepSeek Harness架构解析：从Coding Agent到Agent OS | anionex's blog
DeepSeek Harness架构解析：从Coding Agent到Agent OS | anionex's blog
How Compaction Works in Pi | EARENDIL
How Compaction Works in Pi | EARENDIL
蒸馏、合成数据和 RSI，是怎么连到一起的？
蒸馏、合成数据和 RSI，是怎么连到一起的？
深度剖析 DeepSeek 最新的 Harness DSH：为了自进化这盘醋包了一整盘饺子
深度剖析 DeepSeek 最新的 Harness DSH：为了自进化这盘醋包了一整盘饺子
Designing Apps for Humans and Agents in an Agent-Native Runtime · PocketJS
Designing Apps for Humans and Agents in an Agent-Native Runtime · PocketJS
The human is the loop | Brent Fitzgerald
The human is the loop | Brent Fitzgerald
Practical multi-agent orchestration in Codex
Practical multi-agent orchestration in Codex
Be AI-Native and AI-Proof
Be AI-Native and AI-Proof
Building software factories (with no slop)
Building software factories (with no slop)
AI Engineering Skills Map: Building and Deploying AI Applications
AI Engineering Skills Map: Building and Deploying AI Applications
Harness Engineering for Self-Improvement | Lil'Log
Harness Engineering for Self-Improvement | Lil'Log
别让 LLM 当 JSON 搬运工
别让 LLM 当 JSON 搬运工
The deepseek of DeepSeek Harness: Overengineering or built for self-evolution? | Life Odyssey
The deepseek of DeepSeek Harness: Overengineering or built for self-evolution? | Life Odyssey
异步续延与 Effect 架构的第一性原理
异步续延与 Effect 架构的第一性原理
How Warp builds self-improving agents on Claude | Claude by Anthropic
How Warp builds self-improving agents on Claude | Claude by Anthropic
Introducing SWE-bench Verified | OpenAI
Introducing SWE-bench Verified | OpenAI
Teamwork: When AI Becomes a Research Partner | Google Antigravity Blog
Teamwork: When AI Becomes a Research Partner | Google Antigravity Blog
The end of programming — Paul Dix
The end of programming — Paul Dix
The Harness Is the Thing — Scott Fryxell
The Harness Is the Thing — Scott Fryxell
Inference will eat the world.
Inference will eat the world.
AI Engineering Skills Map: Software engineering fundamentals
AI Engineering Skills Map: Software engineering fundamentals
Running a Software Factory Efficiently at Uber Scale
Running a Software Factory Efficiently at Uber Scale
Agentic Engineering Setup (after 2,000+ hours)
Agentic Engineering Setup (after 2,000+ hours)
Grok Bot for Engineering
Grok Bot for Engineering
基于 Git 的最小知识循环系统
基于 Git 的最小知识循环系统
Hugging Face 事件与未来之路 | OpenAI
Hugging Face 事件与未来之路 | OpenAI
重新设计推理芯片：从 Nvidia GPU 的缺陷到 OpenAI Jalapeño
重新设计推理芯片：从 Nvidia GPU 的缺陷到 OpenAI Jalapeño
Just a rumour of a bug is enough to find a security exploit these days | Anil Madhavapeddy
Just a rumour of a bug is enough to find a security exploit these days | Anil Madhavapeddy
Orchestrate teams of Claude Code sessions - Claude Code Docs
Orchestrate teams of Claude Code sessions - Claude Code Docs
Cybersecurity, in the Agent era
Cybersecurity, in the Agent era
Humans Will Matter.
Humans Will Matter.
How to become a Robotics Engineer in 6 months (RESOURCES)
How to become a Robotics Engineer in 6 months (RESOURCES)
pi-matt-pocock 架构说明
pi-matt-pocock 架构说明
Understanding Deep Learning Requires Rethinking Generalization | susun's blog
Understanding Deep Learning Requires Rethinking Generalization | susun's blog
GitHub - wzhudev/reverse-linear-sync-engine: A reverse engineering of Linear's sync engine. Endorsed by Linear CTO. · GitHub
GitHub - wzhudev/reverse-linear-sync-engine: A reverse engineering of Linear's sync engine. Endorsed by Linear CTO. · GitHub
How to build a shitty robot
How to build a shitty robot
Claude Tag 产品分析 | 白宦成
Claude Tag 产品分析 | 白宦成
从一次性 Agent 到长程控制面 · LoopX
从一次性 Agent 到长程控制面 · LoopX
My PhD Grind (1): All the Friends Who Worked on Theory Have Switched to LLMs / 我的PhD Grind（1）：23年申phd时认识的那群朋友至此都转了LLM | Jinyan Su
My PhD Grind (1): All the Friends Who Worked on Theory Have Switched to LLMs / 我的PhD Grind（1）：23年申phd时认识的那群朋友至此都转了LLM | Jinyan Su
Porting my 1993 Amiga game to Godot, with an LLM reading the 68000 assembly — Babylonian Twins
Porting my 1993 Amiga game to Godot, with an LLM reading the 68000 assembly — Babylonian Twins
Hume Texts Online
Hume Texts Online
The lifecycle of a sharded Postgres query — PlanetScale
The lifecycle of a sharded Postgres query — PlanetScale
Your work might not need the smartest model - Microsoft for Developers
Your work might not need the smartest model - Microsoft for Developers