gguf-quantization

65 / 100 · Review before installing

Check before installing. Some of what it can do is expected: a research skill usually looks things up on the internet.

By KalarisLabs/research-agent-skills (GitHub) · research · 126 stars · scanned at commit 802bbea

GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.

Works with any agent that supports SKILL.md skills.

What this skill can do

Runs commands on your computer

The author does not mention this.

Found in 17 places.

skills/gguf-quantization/SKILL.md, line 45git clone https://github.com/ggml-org/llama.cpp
skills/gguf-quantization/SKILL.md, line 65pip install -r requirements.txt
skills/gguf-quantization/SKILL.md, line 80./llama-quantize model-f16.gguf model-q4_k_m.gguf Q4_K_M

Show 14 more

skills/gguf-quantization/SKILL.md, line 91./llama-cli -m model-q4_k_m.gguf -p "Hello, how are you?"
skills/gguf-quantization/SKILL.md, line 133huggingface-cli download meta-llama/Llama-3.1-8B --local-dir ./llama-3.1-8b
skills/gguf-quantization/SKILL.md, line 151python convert_hf_to_gguf.py ./model --outfile model-f16.gguf
skills/gguf-quantization/SKILL.md, line 179MODEL="llama-3.1-8b-f16.gguf"
skills/gguf-quantization/SKILL.md, line 265./llama-server -m model-q4_k_m.gguf \
skills/gguf-quantization/SKILL.md, line 303make clean && make GGML_METAL=1
skills/gguf-quantization/SKILL.md, line 320make clean && make GGML_CUDA=1
skills/gguf-quantization/SKILL.md, line 333make clean && make
skills/gguf-quantization/SKILL.md, line 353cat > Modelfile << 'EOF'
skills/gguf-quantization/SKILL.md, line 379cp model-q4_k_m.gguf text-generation-webui/models/
skills/gguf-quantization/SKILL.md, line 399./llama-cli -m model.gguf --mmap
skills/gguf-quantization/SKILL.md, line 405./llama-cli -m model.gguf -ngl 20 # Reduce from 35
skills/gguf-quantization/SKILL.md, line 414./llama-imatrix -m model-f16.gguf -f calibration.txt -o model.imatrix
skills/gguf-quantization/references/advanced-usage.md, line 165result = subprocess.run(

Downloads and installs other software

The author does not mention this.

Found in 8 places.

skills/gguf-quantization/SKILL.md, line 58pip install llama-cpp-python
skills/gguf-quantization/SKILL.md, line 65pip install -r requirements.txt
skills/gguf-quantization/references/troubleshooting.md, line 22sudo apt install nvidia-cuda-toolkit

Show 5 more

skills/gguf-quantization/references/troubleshooting.md, line 37pip install cmake scikit-build-core
skills/gguf-quantization/references/troubleshooting.md, line 40CMAKE_ARGS="-DGGML_CUDA=on" pip install llama-cpp-python --force-reinstall --no-cache-dir
skills/gguf-quantization/references/troubleshooting.md, line 43CMAKE_ARGS="-DGGML_METAL=on" pip install llama-cpp-python --force-reinstall --no-cache-dir
skills/gguf-quantization/references/troubleshooting.md, line 55CUDACXX=/usr/local/cuda/bin/nvcc CMAKE_ARGS="-DGGML_CUDA=on" pip install llama-cpp-python
skills/gguf-quantization/references/troubleshooting.md, line 367CMAKE_ARGS="-DGGML_METAL=on" pip install llama-cpp-python --force-reinstall

Can call paid AI services billed to you

The author does not mention this.

Found in 2 places.

skills/gguf-quantization/SKILL.md, line 282from openai import OpenAI
skills/gguf-quantization/references/troubleshooting.md, line 333from openai import OpenAI

Sends data to these websites: api.openai.com

The author does not mention this. Expected for a research skill.

Found in 2 places.

skills/gguf-quantization/SKILL.md, line 282from openai import OpenAI
skills/gguf-quantization/references/troubleshooting.md, line 333from openai import OpenAI

Safer alternatives

Low-risk skills in the same category (research) that look related. Check that one does what you need.

80 Low riskawq-quantizationOrchestra-Research/AI-Research-SKILLs · 14,447 stars

Install

Paste this into your terminal. It downloads exactly the version we scanned, not whatever is newest.

if [ -e ~/.agents/skills/'gguf-quantization' ] || [ -L ~/.agents/skills/'gguf-quantization' ]; then echo 'A skill folder named gguf-quantization already exists — not replacing it' >&2; false; else d=$(mktemp -d) && git clone --quiet --filter=blob:none --no-checkout 'https://github.com/KalarisLabs/research-agent-skills' "$d" && git -C "$d" checkout --quiet '802bbea44e551abac24d6856f83fc55149bbdba3' -- 'skills/gguf-quantization' && mkdir -p ~/.agents/skills && cp -R "$d"/'skills/gguf-quantization' ~/.agents/skills/'gguf-quantization' && rm -rf "$d"; fi

Codex reads ~/.agents/skills directly. For Claude Code, Cursor, Copilot, Gemini CLI, Antigravity or OpenCode, also link it. For Claude Code: ln -s ~/.agents/skills/'gguf-quantization' ~/.claude/skills/'gguf-quantization' (other agents: ~/.cursor/skills, ~/.copilot/skills, ~/.gemini/skills, ~/.gemini/config/skills, ~/.config/opencode/skills). Or install through the finder skill, which does both.

Badge for your README

The badge for KalarisLabs/research-agent-skills shows the score of its lowest-scoring skill.

SkillTrust badge for KalarisLabs/research-agent-skills

[![SkillTrust](https://zalavijaysinh.github.io/skilltrust/badges/KalarisLabs/research-agent-skills.svg)](https://zalavijaysinh.github.io/skilltrust/repos/KalarisLabs/research-agent-skills/)

Score history

Date	Version	Score
2026-11-13	`802bbea`	65
2026-11-07	`29fe36d`	100