Agent skill

Hf Architecture Tikz

by yzlnew in yzlnew/infra-skills

Draw Sebastian-Raschka-gallery-style TikZ architecture diagrams for any HuggingFace decoder-only LLM, with per-block parameter formulas and concrete numbers.

No licenceAuto-check passedDocuments & Office

Install Hf Architecture Tikz

skills CLI
$ npx skills add yzlnew/infra-skills --skill hf-architecture-tikz -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install yzlnew/infra-skills hf-architecture-tikz --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/yzlnew/infra-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/hf-architecture-tikz .claude/skills/hf-architecture-tikz && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
hf-architecture-tikz
GitHub stars
149
Token cost
~2k tokens
SKILL.md length
783 words
Files
12 (incl. scripts, references)
Skills in repo
8
Repo updated
First seen
Licence
None found

At a glance

Draw Sebastian-Raschka-gallery-style TikZ architecture diagrams for any HuggingFace decoder-only LLM, with per-block parameter formulas and concrete numbers.

  • Works in 5 steps: Acquire config. extract_arch.py tries… → Detect architecture family. Pure… → Compute parameter counts. Closed-form… → …
  • The user asks to visualize / diagram / illustrate a transformer
  • SKILL.md covers When to use, Quick start, Workflow and Architecture family detection, plus 5 more sections
  • Runs Python and Shell scripts from its folder; calls uv and bash

What it does

Hf Architecture Tikz is an agent skill from yzlnew/infra-skills. Draw Sebastian-Raschka-gallery-style TikZ architecture diagrams for any HuggingFace decoder-only LLM, with per-block parameter formulas and concrete numbers. Supports MHA, GQA, MLA, DeepSeek-V4-Flash (Hyper-Connections + Sparse Attention with learned indexer), dense and MoE FFNs (incl. hash routing), and MTP heads. Use when the user asks to visualize / diagram / illustrate a transformer or LLM architecture (DeepSeek, Qwen, Llama, Mistral, gpt-oss, etc.), wants a Raschka-style figure, or wants a TikZ/LaTeX…

Its SKILL.md is about 2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 16 other files, including scripts and reference files (for example `examples/deepseek-v4-flash/arch.json`, `references/architecture_families.md` and `references/param_formulas.md`).

It sits in Documents & Office, covering Diagrams, Model hubs and datasets and LaTeX. It works with DeepSeek, Hugging Face, Mistral AI and Qwen. The repository describes itself as: A collection of specialized agent skills for AI infrastructure development, enabling Claude Code to write, optimize, and debug high-performance systems.

When your agent uses it

  • The user asks to visualize / diagram / illustrate a transformer
  • LLM architecture (DeepSeek
  • Wants a Raschka-style figure
  • Wants a TikZ/LaTeX rendering of an HF model

Example prompts

  • “/hf-architecture-tikz”

Requirements

  • Python 3
  • A Bash shell

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

  1. Acquire config. extract_arch.py tries transformers.AutoConfig first; if the installed transformers doesn't recognize the model_type (e.g…
  2. Detect architecture family. Pure config-field rules — see references/architecture_families.md. The script labels the model with a family…
  3. Compute parameter counts. Closed-form formulas keyed by family — see references/param_formulas.md. The script (not Claude) does the…
  4. Assemble TikZ. render_tikz.py reads arch.json plus templates/anthropic.tex.j2 (Jinja2 template — all block macros are inlined for shared…
  5. Compile. bash scripts/compile.sh out.tex runs xelatex ×2 (TikZ fit/positioning needs a second pass) then pdftocairo -png -r 300…

What it can do on your machine

Read from SKILL.md and the folder at commit f3a8d7d. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 3 files in scripts/ (Python and Shell), which the agent can run.

    Shell commands in SKILL.md call:

    • uv
    • bash

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Links to these hosts (documentation or services it may open):

    • sebastianraschka.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Hf Architecture Tikz loads about 2k tokens when it runs, and up to ~6.1k if it reads all its reference files. Until then it costs about 139 tokens; SKILL.md has 783 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~139
When it runs · the whole SKILL.md, loaded when a task matches
~2k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~6.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

Without a licence we can't republish the file, so here is its outline and opening line. It has 783 words (~1,992 tokens).

“Generate a publication-quality vertical architecture diagram (in the style of Sebastian Raschka's LLM Architecture Gallery) for any HuggingFace decoder-only LLM. The diagram annotates every sub-block with its parameter-count formula and the concrete number for the loaded config.”

— opening of SKILL.md by yzlnew
name
hf-architecture-tikz

Read the full SKILL.md on GitHub

Files

SKILL.md and 11 other files (scripts, references) in hf-architecture-tikz of yzlnew/infra-skills.

  • SKILL.md
  • examples/deepseek-v4-flash/arch.json
  • examples/deepseek-v4-flash/deepseek-v4-flash.pdf
  • examples/deepseek-v4-flash/deepseek-v4-flash.png
  • examples/deepseek-v4-flash/deepseek-v4-flash.tex
  • references/architecture_families.md
  • references/param_formulas.md
  • references/raschka_style_guide.md
  • scripts/compile.sh
  • scripts/extract_arch.py
  • scripts/render_tikz.py
  • templates/anthropic.tex.j2

Open the folder on GitHubat commit f3a8d7d

Compare with similar skills

Hf Architecture Tikz next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Hf Architecture Tikz compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Hf Architecture Tikz this skillyzlnew/infra-skills149—~2kAutomated safety check: PassNone
Write Leetcuda Bookxlite-dev/LeetCUDA12k—~2.1kAutomated safety check: PassGPL-3.0
Add Modelguoqingbao/xinfer334—~4.2kAutomated safety check: NotesMIT
Setup Benchmark Inputsmlc-ai/pith-train355—~399Automated safety check: PassApache-2.0
Quark Torch Ptqamd/Quark182—~2.3kAutomated safety check: PassMIT
Tikz Diagrams Guidewentorai/research-plugins2981 repos~1.4kAutomated safety check: PassMIT

Similar skills

  • Write Leetcuda Book

    xlite-dev/LeetCUDA

    把 LeetCUDA kernels/interview 教学源码(base/sgemv/sgemm/hgemm/flashattn/ffpaattn/fp8gemm.cuh/fp4gemm.cuh)+ ffpa-attn CuTe sm120 源码(csrc/cuffpa/cute fp8/fp4,commit 861d75e)写成中文 CUDA 技术书(6 Part 38 章 + CuTe…

    12k GitHub stars~2.1k tokensUpdated today
    Documents & OfficeAuto-check passed
  • Add Model

    guoqingbao/xinfer

    Adapt and port new LLM model architectures to this xinfer project.

    334 GitHub stars~4.2k tokensUpdated 1 mo ago
    AI & LLM EngineeringAuto-check: notes
  • Setup Benchmark Inputs

    mlc-ai/pith-train

    Set up the minimal set of artifacts (tokenized DCLM corpus shard + released HuggingFace checkpoint converted to DCP) required to benchmark, profile, or regression-test a MoE model in PithTrain.

    355 GitHub stars~399 tokensUpdated yesterday
    AI & LLM EngineeringAuto-check passed
  • Runs an end-to-end AMD Quark post-training quantization workflow for PyTorch / Hugging Face LLMs: inspect a Hub or local model, choose a quantization plan, create reproducible artifacts, request…

    182 GitHub stars~2.3k tokensUpdated 13 days ago
    AI & LLM EngineeringAuto-check passed
  • Tikz Diagrams Guide

    wentorai/research-plugins

    Create publication-quality scientific diagrams with TikZ in LaTeX

    298 GitHub starsUsed in 1 repo~1.4k tokens
    Documents & OfficeAuto-check passed
  • Open Weights

    ericrisco/rsc-harness

    A skill your agent uses when choosing an open-weight LLM and clearing it for use — which family and size fit the task, the hardware and the budget, and above all whether the license permits shipping.

    180 GitHub stars~4.1k tokensUpdated yesterday
    AI & LLM EngineeringAuto-check passed

More from yzlnew/infra-skills

All 8 skills in this repo
  • Megatron Memory Estimator

    yzlnew/infra-skills

    Estimate GPU memory usage for Megatron-based MoE (Mixture of Experts) and dense models.

    149 GitHub stars~2.2k tokensUpdated 3 mo ago
    Auto-check passed
  • HTML Flowchart Anthropic

    yzlnew/infra-skills

    Create and revise pure HTML/CSS flowcharts using an Anthropic-inspired design language.

    149 GitHub stars~1.5k tokensUpdated 3 mo ago
    Auto-check passed
  • Openai Dotcom Viz

    yzlnew/infra-skills

    Build figures in OpenAI's blog / research / system-card "dotcom" visual style — both (a) bar charts (monochrome bars with a darker same-hue stroke, rounded corners, a black y-axis with outward ticks…

    149 GitHub stars~1.3k tokensUpdated 3 mo ago
    Auto-check passed
  • Slime User

    yzlnew/infra-skills

    Guide for using SLIME (LLM post-training framework for RL Scaling).

    149 GitHub stars~3.2k tokensUpdated 3 mo ago
    Auto-check passed
  • Tilelang Developer

    yzlnew/infra-skills

    Write, optimize, and debug high-performance AI compute kernels using TileLang (a Python DSL for GPU programming).

    149 GitHub stars~2.4k tokensUpdated 3 mo ago
    Auto-check passed
  • Material You Slides

    yzlnew/infra-skills

    Create presentation slides using Material You (Material Design 3) style.

    149 GitHub stars~3.7k tokensUpdated 3 mo ago
    Auto-check passed

Questions about Hf Architecture Tikz

What does Hf Architecture Tikz do?

Draw Sebastian-Raschka-gallery-style TikZ architecture diagrams for any HuggingFace decoder-only LLM, with per-block parameter formulas and concrete numbers. Hf Architecture Tikz is an agent skill from yzlnew/infra-skills. Draw Sebastian-Raschka-gallery-style TikZ architecture diagrams for any HuggingFace decoder-only LLM, with per-block parameter formulas and concrete numbers.

When should I use Hf Architecture Tikz?

Hf Architecture Tikz fits situations like: the user asks to visualize / diagram / illustrate a transformer; LLM architecture (DeepSeek; wants a Raschka-style figure; wants a TikZ/LaTeX rendering of an HF model.

How do I install Hf Architecture Tikz in Claude Code?

Run `npx skills add yzlnew/infra-skills --skill hf-architecture-tikz -a claude-code`. Or copy the skill folder (hf-architecture-tikz in yzlnew/infra-skills) into .claude/skills/hf-architecture-tikz in your project. Claude Code loads it when a task matches its description.

How do I install Hf Architecture Tikz in Codex?

Run `npx skills add yzlnew/infra-skills --skill hf-architecture-tikz -a codex`. Or copy the skill folder (hf-architecture-tikz in yzlnew/infra-skills) into .agents/skills/hf-architecture-tikz in your project. Codex loads it when a task matches its description.

Can I use Hf Architecture Tikz in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add yzlnew/infra-skills --skill hf-architecture-tikz -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/hf-architecture-tikz, .gemini/skills/hf-architecture-tikz, .github/skills/hf-architecture-tikz and .opencode/skills/hf-architecture-tikz in your project.

What does Hf Architecture Tikz need to run?

Going by SKILL.md and its folder, Hf Architecture Tikz needs Python and a shell for the scripts in its folder and the command-line tools its instructions call (uv and bash). Our summary lists: Python 3; A Bash shell.

Does Hf Architecture Tikz access the network?

SKILL.md names 1 domain. As links in the text: sebastianraschka.com. This is read from the text; nothing was executed.

Is Hf Architecture Tikz safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Hf Architecture Tikz use?

No licence was found for Hf Architecture Tikz or its repository. Without one, default copyright applies: ask the author before reusing or redistributing it.

How many tokens does Hf Architecture Tikz use?

About 2k tokens (SKILL.md is roughly 8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 4.1k tokens, read only when the agent opens those files.

What are the alternatives to Hf Architecture Tikz?

Skills that share tags, products or a category with Hf Architecture Tikz: Write Leetcuda Book (xlite-dev/LeetCUDA, 12k stars), Add Model (guoqingbao/xinfer, 334 stars), Setup Benchmark Inputs (mlc-ai/pith-train, 355 stars) and Quark Torch Ptq (amd/Quark, 182 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Hf Architecture Tikz?

yzlnew (a GitHub user) maintains it in yzlnew/infra-skills, which has 149 GitHub stars. The repository holds 8 skills in this directory. The repository was last updated on July 9, 2026.

Source: yzlnew/infra-skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.