unsloth

Unsloth: 2-5x faster LoRA/QLoRA fine-tuning, less VRAM.

Quality

40%

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

Fix and improve this skill with Tessl

tessl review fix ./optional-skills/mlops/training/unsloth/SKILL.md

Quality

Content

22%

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

The body is lean auto-generated scaffolding that adds little actionable value: no code, no sequenced workflow, and only one of several bundled reference files is surfaced. It reads as an unfinished template rather than a working skill.

Suggestions

Replace the placeholder Quick Reference with one or two copy-paste-ready Unsloth fine-tuning code snippets to lift actionability from 1.

Add a sequenced workflow (load model -> apply LoRA -> train -> save) with a verification checkpoint (e.g., checking the trainer output/loss) to address workflow_clarity.

Remove filler lines ('patterns will be added as you use the skill', 'Add helper scripts here') and list each bundled reference file with a one-line purpose so navigation is clear.

Dimension	Reasoning	Score
Conciseness	The body is short and does not pad with concepts Claude already knows, but much of it is placeholder filler ('Quick reference patterns will be added as you use the skill', 'Add helper scripts here') whose tokens do not earn their place, so it does not reach the lean-and-efficient anchor at 3.	2 / 3
Actionability	There is no executable code, command, or concrete example anywhere in the body; guidance like 'Use `view` to read specific reference files' is abstract direction rather than instruction, matching the 'vague or abstract; no concrete code/commands' anchor.	1 / 3
Workflow Clarity	No multi-step process is sequenced and there are no validation checkpoints; the 'When to Use' and 'Working with This Skill' sections are generic categories, not an ordered workflow, matching the 'steps unclear or missing' anchor.	1 / 3
Progressive Disclosure	The body signals one real one-level-deep reference (llms-txt.md, which exists in references/), but it omits the other large bundled files (llms-full.md, llms.md, index.md) and offers no navigation among them, matching the 'references present but not clearly signaled' anchor.	2 / 3
	Total	6 / 12 Passed

Description

57%

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 a concise, distinct tagline that clearly identifies the Unsloth niche, but it lacks an explicit 'Use when...' trigger clause and concrete action verbs, capping both completeness and specificity at 2.

Suggestions

Add a 'Use when...' clause naming the natural trigger phrases (e.g., 'Use when the user wants to fine-tune or train Llama, Mistral, Gemma, or Qwen models with LoRA/QLoRA, or asks for faster/low-VRAM training').

Lead with concrete action verbs instead of a tagline (e.g., 'Fine-tunes models with LoRA/QLoRA at 2-5x speed and lower VRAM') to lift specificity to 3.

Include common user phrasings like 'train', 'fine-tune', and 'lora training' to broaden trigger-term coverage.

Dimension	Reasoning	Score
Specificity	Names a concrete domain ('LoRA/QLoRA fine-tuning') and quantified outcomes ('2-5x faster', 'less VRAM'), but it is a tagline with no action verbs, so it falls short of the 'multiple specific concrete actions' anchor at 3.	2 / 3
Completeness	It states what the skill does (faster LoRA/QLoRA fine-tuning, less VRAM) but has no 'Use when...' clause or equivalent trigger guidance; per the judging guidelines, a missing explicit trigger caps completeness at 2.	2 / 3
Trigger Term Quality	Includes relevant keywords a user might say ('Unsloth', 'fine-tuning', 'LoRA/QLoRA', 'VRAM') but misses common variations like 'train', 'fine-tune a model', or 'lora training', matching the 'some relevant keywords but missing common variations' anchor.	2 / 3
Distinctiveness Conflict Risk	Naming the specific library 'Unsloth' with its niche (LoRA/QLoRA fine-tuning acceleration) gives it a clear, distinct trigger unlikely to fire for unrelated skills.	3 / 3
	Total	9 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: NousResearch/hermes-agent
Path: optional-skills/mlops/training/unsloth/SKILL.md
Commit: 4c9628e

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.