llama-factory

Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support

Quality

55%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./backend/cli/skills/ml-training/llama-factory/SKILL.md

Quality

Content

40%Weight 40%Scale 1-5

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

The body functions as a thin overview delegating everything to reference files, but it offers almost no actionable guidance, no executable examples, and no real workflow for the core fine-tuning task. Progressive disclosure is reasonable though vaguely labeled.

Suggestions

Replace the empty 'Quick Reference / Common Patterns' placeholder with 2-3 concrete, copy-paste-ready fine-tuning command examples (e.g. a WebUI launch command and a CLI LoRA training command).

Add a short sequenced workflow for a typical fine-tuning run (prepare dataset -> train -> validate/evaluate -> merge/export) with an explicit validation checkpoint before merging.

Replace generic reference labels ('Advanced documentation') with topic-specific descriptions so each reference file's contents are immediately discoverable.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient but contains boilerplate filler and placeholder sections ('Quick reference patterns will be added as you use the skill', empty scripts/assets invitations) that could be trimmed.	3 / 5
Actionability	Provides only high-level navigation hints to reference files with no executable code, concrete commands, or real examples; the Quick Reference section is an empty placeholder.	2 / 5
Workflow Clarity	No sequenced workflow for actually fine-tuning a model is present, and the risky multi-step training operation has no validation checkpoints; only a generic routing scheme and an unrelated 'Updating' procedure exist.	2 / 5
Progressive Disclosure	SKILL.md is an overview pointing to real one-level-deep reference files (advanced.md, getting_started.md, other.md, index.md), but the reference descriptions are generic rather than keyed to specific sub-topics.	4 / 5
	Total	11 / 20 Passed

Description

70%Weight 40%Scale 1-5

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 conveys a clear, distinct niche and several concrete capabilities but omits any explicit 'when to use' trigger guidance, which the rubric caps at completeness 3. It is concise and specific but lacks natural trigger phrasings.

Suggestions

Add an explicit 'Use when...' clause stating concrete triggers, e.g. 'Use when the user wants to fine-tune an LLM, run LoRA/QLoRA training, or use LLaMA-Factory's WebUI.'

Soften the jargon with a natural phrasing users would actually say, such as 'fine-tune a model' or 'LoRA training', alongside the technical terms.

Reframe feature lists as concrete actions (train, merge LoRA weights, evaluate, export) rather than only naming supported capabilities.

Dimension	Reasoning	Score
Specificity	Lists several concrete capabilities (WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal) but frames them as supported features rather than discrete actions, leaving minor coverage gaps.	4 / 5
Completeness	Provides a clear 'what' (fine-tuning LLMs with LLaMA-Factory) but has no 'Use when...' or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric.	3 / 5
Trigger Term Quality	Includes real keywords (fine-tuning LLMs, LLaMA-Factory, WebUI, QLoRA, multimodal) but is jargon-heavy and missing natural user phrasings like 'fine-tune a model' or 'LoRA training'.	4 / 5
Distinctiveness Conflict Risk	Names the specific tool (LLaMA-Factory) alongside distinctive features like WebUI no-code and QLoRA, giving it a clear niche with minimal conflict risk.	5 / 5
	Total	16 / 20 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-training/llama-factory/SKILL.md
Commit: fcd2e11

Reviewed: 2 days 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.