gguf-quantization

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.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

72%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with executable code throughout and good progressive disclosure via real reference files. It is somewhat padded with concept re-explanation and lacks explicit validation checkpoints in its batch/destructive workflows.

Suggestions

Trim the GGUF definition restatement and the 'Key advantages'/'Use alternatives' concept explanations; assume Claude knows what GGUF and the surrounding ecosystem are.

Add an explicit validation/verification step to the batch quantization workflow (e.g., test-load each output with llama-cli and check size/perplexity before declaring success).

Add a verify checkpoint after model conversion (e.g., load the FP16 GGUF and sanity-check before quantizing) to close the feedback loop on irreversible operations.

Dimension	Reasoning	Score
Conciseness	Mostly efficient executable code, but the intro re-expands the GGUF acronym/concept and the 'Key advantages'/'Use alternatives' sections add conceptual padding Claude already knows.	2 / 3
Actionability	Provides copy-paste-ready shell commands and complete runnable Python across conversion, quantization, inference, server, and integration sections.	3 / 3
Workflow Clarity	Conversion workflows are sequenced, but batch/irreversible operations like the Workflow 3 multi-quant loop and model conversion lack explicit validation or verification checkpoints.	2 / 3
Progressive Disclosure	Body serves as an overview with two well-signaled, one-level-deep references (references/advanced-usage.md, references/troubleshooting.md) that exist as real files.	3 / 3
	Total	10 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is concise, specific, and complete, clearly stating both the capability and explicit use-when triggers in third person. It carves out a distinct niche with natural trigger terms.

Dimension	Reasoning	Score
Specificity	Names concrete capabilities — 'GGUF format and llama.cpp quantization', 'efficient CPU/GPU inference', 'flexible quantization from 2-8 bit' — rather than vague language.	3 / 3
Completeness	Explicitly answers both what it does (GGUF/llama.cpp quantization for CPU/GPU inference) and when to use it via a clear 'Use when...' clause with explicit triggers.	3 / 3
Trigger Term Quality	Covers natural terms users would say — 'GGUF', 'quantization', 'consumer hardware', 'Apple Silicon', '2-8 bit' — with good coverage of common phrasings.	3 / 3
Distinctiveness Conflict Risk	Niche is clearly scoped to GGUF/llama.cpp CPU-GPU inference, distinct from alternatives like AWQ/GPTQ, so unlikely to trigger for the wrong skill.	3 / 3
	Total	12 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/ml-inference/gguf/SKILL.md
Commit: 52845c3

Reviewed: about 14 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.