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 ./skills/fine-tuning/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 is a generic auto-generated scaffold: it organizes reference files reasonably well but provides almost no actionable guidance, no executable examples, and no real fine-tuning workflow or validation steps.

Suggestions

Replace the empty "Quick reference patterns will be added as you use the skill" with 2-3 concrete, copy-pasteable fine-tuning examples (e.g. a WebUI launch command and a QLoRA YAML config snippet).

Add a step-by-step fine-tuning workflow with explicit validation checkpoints (dataset format check, dry-run, loss/eval monitoring) since training is a batch/destructive operation — without it workflow_clarity is capped.

Remove placeholder sections ("Add helper scripts here", "Add templates, boilerplate, or example projects here") and the auto-generation meta-notes, and link reference files with markdown links plus one-line topical descriptions instead of generic labels.

Dimension	Reasoning	Score
Conciseness	The body is short and does not over-explain concepts Claude already knows, but it carries unnecessary placeholder/meta padding ("Quick reference patterns will be added as you use the skill", "Add helper scripts here", auto-generation notes) that does not earn its tokens.	3 / 5
Actionability	Only minimal concrete guidance is present — pointers like "Start with the getting_started... reference files" and "Use `view` to read specific reference files" — with no executable code, commands, or specific steps; below 3 because the Quick Reference section is explicitly empty.	2 / 5
Workflow Clarity	No workflow exists for the skill's actual task (fine-tuning); the only sequence is a trivial 2-step "Updating" procedure (re-run scraper, rebuild) with no validation, and fine-tuning is a batch/destructive operation that lacks any checkpoints.	2 / 5
Progressive Disclosure	Reference files are listed one level deep in a dedicated "Reference Files" section with a navigation cue ("Use `view` to read specific reference files"), and all referenced files exist; not a 5 because descriptions are generic placeholders, files are named in backticks rather than linked, and index.md is not mentioned.	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 is specific, concrete, and distinctive, with strong natural trigger terms, but it lacks any explicit "Use when..." trigger guidance, which caps its completeness at 3.

Suggestions

Add an explicit trigger clause, e.g. "Use when fine-tuning or QLoRA-training LLMs with LLaMA-Factory, configuring WebUI training runs, or working with multimodal model training."

Lead with the primary action verb (e.g. "Fine-tunes LLMs via LLaMA-Factory") in third person rather than the noun phrase "Expert guidance for...".

Add natural synonyms a user might say (LoRA, SFT, instruction tuning, RLHF) to broaden trigger coverage toward a 5.

Dimension	Reasoning	Score
Specificity	Lists several specific concrete capabilities — "fine-tuning LLMs", "WebUI no-code", "2/3/4/5/6/8-bit QLoRA", "multimodal support", "100+ models" — with only minor coverage gaps; slightly below 5 because most items are features/techniques rather than distinct actions.	4 / 5
Completeness	The "what" is clear ("Expert guidance for fine-tuning LLMs with LLaMA-Factory..."), but there is no "Use when..." clause or equivalent explicit trigger guidance, which per the rubric caps completeness at 3.	3 / 5
Trigger Term Quality	Includes natural terms users would say ("fine-tuning LLMs", "LLaMA-Factory", "QLoRA", "WebUI no-code", "multimodal") with good coverage; a few natural synonyms (LoRA, SFT, instruction tuning) are missing, so not a 5.	4 / 5
Distinctiveness Conflict Risk	The named tool "LLaMA-Factory" plus "QLoRA" and "WebUI no-code fine-tuning" carve a clear, distinctive niche with minimal overlap risk against other skills.	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: OpenLAIR/dr-claw
Path: skills/fine-tuning/llama-factory/SKILL.md
Commit: 4b35756

Reviewed: 1 day 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.