axolotl

Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support

Quality

66%

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 ./03-fine-tuning/axolotl/SKILL.md

Quality

Content

50%

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

The body is an auto-generated reference catalog with some useful concrete snippets, but it is padded with meta sections, contains broken/placeholder references, and lacks a real sequenced workflow. Tightening the patterns and fixing navigation would lift every dimension.

Suggestions

Remove or trim the Notes, Resources, and 'Working with This Skill' meta sections that restate information Claude already has.

Fix or remove references to nonexistent files (getting_started, tutorials, guides) and delete the empty scripts/assets placeholder sections, or actually populate them.

Replace the bare-token code blocks (Patterns 3 and 6) with complete executable examples, and add a short end-to-end fine-tuning workflow with a validation/eval checkpoint.

Dimension	Reasoning	Score
Conciseness	The Quick Reference is mostly efficient, but the Notes, Resources, and 'Working with This Skill' sections repeat meta filler ('generated from official documentation', 'Add helper scripts here') that Claude does not need.	2 / 3
Actionability	Some concrete guidance exists (FSDP YAML block, NCCL test command), but several patterns are bare tokens in code blocks ('context_parallel_size', 'integrations') and the API examples are raw signatures missing usage context.	2 / 3
Workflow Clarity	No sequenced workflow with validation checkpoints exists for the core fine-tuning task; only a minimal two-step 'Updating' sequence is present, so checkpoints and feedback loops are absent.	2 / 3
Progressive Disclosure	References are one level deep and listed with descriptions, but the body cites nonexistent files ('getting_started', 'tutorials', 'guides') and the scripts/ and assets/ sections are empty placeholders.	2 / 3
	Total	8 / 12 Passed

Description

82%

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 and trigger-rich with a clear niche, but it omits an explicit 'when to use' clause, which caps completeness at 2. Adding a 'Use when...' sentence would round it out.

Suggestions

Append an explicit trigger clause such as 'Use when fine-tuning or instruction-tuning HuggingFace models with Axolotl, configuring YAML training configs, or setting up LoRA/QLoRA/DPO/GRPO runs.'

Clarify the core action verb up front ('Fine-tune and instruction-tune LLMs with Axolotl') so the primary capability reads as an action rather than 'guidance'.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities ('YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support') rather than vague language, matching the 'lists multiple specific concrete actions' anchor.	3 / 3
Completeness	It states what the skill does ('Expert guidance for fine-tuning LLMs with Axolotl...') but lacks any 'Use when...' clause or equivalent explicit trigger guidance, so completeness is capped at 2 per the judging guidelines.	2 / 3
Trigger Term Quality	Includes natural terms a user would say ('fine-tuning LLMs', 'Axolotl', 'LoRA', 'QLoRA', 'DPO', 'GRPO', 'YAML configs'), giving good coverage of likely trigger vocabulary.	3 / 3
Distinctiveness Conflict Risk	The Axolotl-specific niche plus method-level qualifiers make it clearly distinguishable and unlikely to trigger for the wrong skill.	3 / 3
	Total	11 / 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: Orchestra-Research/AI-Research-SKILLs
Path: 03-fine-tuning/axolotl/SKILL.md
Commit: 773a529

Reviewed: about 13 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.