agentreadme

Report · marked 26 Aug 2026

rasbt/LLMs-from-scratch

An agent will struggle here, mostly for fixable reasons.

Jupyter Notebook · 103,780 stars · 353 files · branch main

54 /100 C+
Instructions 2/27Whether the repo tells an agent how to behave before it starts guessing. Losing marks on agent instruction file and instruction quality.Setup 3/14Whether an agent can install the project and get it running without a human. Losing marks on deterministic install and discoverable commands.Verification loop 22/25Whether an agent can check its own work. This is the category that most decides whether agent output is trustworthy. Losing marks on static type checking.Context economy 20/20Whether the repo fits in a context window, or fights it. Full marks.Navigability 5/10How quickly anything new can orient itself. Losing marks on written documentation and license.

Higher is better. Accent marks what is costing you most.

Fix these first

01
Agent instruction file
Add an AGENTS.md at the root: what the project is, how to install, how to run, how to test, and the two or three conventions a newcomer always gets wrong.
02
Instruction quality
Instructions earn their keep by naming exact commands. 'Run the tests with `pnpm test`' beats three paragraphs of philosophy.
03
Discoverable commands
Declare the handful of commands that matter in package.json scripts, a Makefile, or a justfile. Naming them turns guesswork into a lookup.

The full marking

Every deduction below names the file or setting it came from.

Instructions 2/27 F

Whether the repo tells an agent how to behave before it starts guessing.

✗Agent instruction file 0/12
None found. Every agent that opens this repo starts from zero.
Add an AGENTS.md at the root: what the project is, how to install, how to run, how to test, and the two or three conventions a newcomer always gets wrong.
Looked for AGENTS.md, CLAUDE.md, .github/copilot-instructions.md, .cursorrules
✗Instruction quality 0/12
Nothing to judge, since there are no instructions.
Instructions earn their keep by naming exact commands. 'Run the tests with `pnpm test`' beats three paragraphs of philosophy.
✗README as an entry point 2/3
README exists but never explains how to get the thing running.
Give the README an Install and a Usage heading with real commands under each. Agents pattern-match on those headings.
Setup 3/14 F

Whether an agent can install the project and get it running without a human.

✗Deterministic install 3/6
Only requirements.txt, which pins loosely at best.
Commit the lockfile your package manager generates. Without it, an agent's install and your install are different builds, and 'works on my machine' becomes unfalsifiable.
✗Discoverable commands 0/6
No declared commands. An agent has to infer how to build and run this from the file tree.
Declare the handful of commands that matter in package.json scripts, a Makefile, or a justfile. Naming them turns guesswork into a lookup.
–Environment config n/a
Not applicable — this reads as a library, with no environment to configure.
✗Pinned runtime version 0/2
No pinned runtime version.
Add a .nvmrc, .python-version, or equivalent so the agent's toolchain matches yours.
–Reproducible environment n/a
Not applicable — a library doesn't need a container to be worked on.
Verification loop 22/25 A

Whether an agent can check its own work. This is the category that most decides whether agent output is trustworthy.

✓Tests exist 8/8
39 test files against 149 source files.
e.g. .github/scripts/test_check_links.py, ch03/02_bonus_efficient-multihead-attention/tests/test_mha_implementations.py, ch04/09_dsa/test_dsa.py
✓Test command is discoverable 7/7
An agent can find and run `pytest`.
✓Continuous integration 4/4
12 GitHub Actions workflows define what "passing" means.
✓Lint and format rules 3/3
pyproject.toml encodes the house style.
✗Static type checking 0/3
No static type checking.
A type checker gives an agent an error message instead of a runtime surprise. It's the second-fastest feedback loop after the compiler.
Context economy 20/20 A+

Whether the repo fits in a context window, or fights it.

✓No committed build output 6/6
No generated directories committed.
✓.gitignore hygiene 3/3
.gitignore covers 233 patterns.
✓File sizes fit in context 6/6
No source file is large enough to crowd out a context window.
✓Repository weight 5/5
About 16.9MB checked out, which is comfortable.

Here is your AGENTS.md

Drafted from what is actually in this repository: the install command from your lockfile, the commands you already declare, your real directory layout. Anything marked TODO needs a person. Save it at the root as AGENTS.md.

AGENTS.md — drafted for rasbt/LLMs-from-scratch 
# AGENTS.md

Implement a ChatGPT-like LLM in PyTorch from scratch, step by step

## Setup

```
pip install -r requirements.txt
```

## Commands

```
ruff check .   # lint
```

## Layout

- `ch05/`       41 source files
- `pkg/`        37 source files
- `ch04/`       32 source files
- `ch06/`       16 source files
- `ch07/`       9 source files
- `appendix-A/` 2 source files

## Conventions

- Tests live alongside the code they cover, following `.github/scripts/test_check_links.py`.
- CI defines what passing means. See `.github/workflows/basic-tests-latest-python.yml`, and keep it green.
- TODO: add the two or three conventions a newcomer always gets wrong here.

## Gotchas

- No lockfile is committed, so an install here may not match what CI produced.

---

Drafted by agentreadme.com from what is in this repository. Everything marked TODO
needs a human. Check it in as AGENTS.md at the root.

Open the raw markdown  or  curl -o AGENTS.md agentreadme.com/draft/rasbt/LLMs-from-scratch.md

Show the mark

The badge re-checks daily, so it keeps up as the repository changes. Use mark again to force it now.

agent ready 54 out of 100

[![agent ready](https://agentreadme.com/badge/rasbt/LLMs-from-scratch.svg)](https://agentreadme.com/rasbt/LLMs-from-scratch)

Other Jupyter Notebook repositories, marked

jackfrued/Python-100-Days 22 F microsoft/generative-ai-for-beginners 88 A microsoft/ML-For-Beginners 76 B+ openai/openai-cookbook 68 B CompVis/stable-diffusion 32 D microsoft/ai-agents-for-beginners 66 B

All Jupyter Notebook repositories

Think this mark is wrong?

Every deduction above names the file it came from, so this can be settled by looking. If a check missed something, that is a rule worth fixing.

Open an issue, already filled in