stable-diffusion

Text-to-image generation, inpainting, and img2img.

Quality

61%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./optional-skills/mlops/stable-diffusion/SKILL.md

Quality

Content

67%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 well-structured, highly actionable, and supported by genuine one-level-deep reference files. Its weaknesses are moderate verbosity (architecture explanations Claude doesn't need, duplicated memory guidance) and one non-executable code fragment, which keep it just above the rubric's midpoint rather than at the exemplar level.

Suggestions

Trim or remove the 'Architecture overview' diagrams and the duplicated memory-optimization content (keep it once, or defer to references/troubleshooting.md), cutting the body toward a true overview.

Define or replace `get_canny_image()` in the ControlNet example with a complete snippet (e.g., using cv2.Canny and a PIL passthrough) so all code is executable.

Move the ControlNet, LoRA, and model-variant detail sections into references and leave one short example plus pointers in SKILL.md to tighten progressive disclosure.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient code and tables, but the 'Architecture overview' section spends ~20 lines diagramming diffusers internals (pipeline/UNet/VAE/scheduler, inference flow) that Claude already knows, and memory-optimization guidance appears both in its own section and again under 'Common issues' (CUDA out of memory).	3 / 5
Actionability	Nearly all code blocks are complete, copy-paste-ready with real model IDs and parameters, but the ControlNet example calls `get_canny_image(input_image)`, a helper that is never defined anywhere in the skill, which is exactly a 'concrete code with minor gaps' case.	4 / 5
Workflow Clarity	Sections follow a clear sequence (install → quick start → concepts → tasks → issues) and the 'Common issues' section provides error-recovery guidance (OOM → offload/slicing options, black images → checker/dtype fixes), but there are no explicit verification checkpoints; since generation is non-destructive, this lands at 'clear sequence with minor validation gaps'.	4 / 5
Progressive Disclosure	Two real one-level-deep reference files (references/advanced-usage.md, references/troubleshooting.md) are listed in a labeled References section, but the 523-line body is a comprehensive guide rather than an overview, and its 'Common issues' section overlaps with troubleshooting.md instead of deferring to it.	4 / 5
	Total	15 / 20 Passed

Description

55%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 concise and names three concrete, distinct capabilities in third-person voice, but it lacks any 'when to use' trigger guidance and misses key natural trigger terms like "Stable Diffusion" and "image generation". It is serviceable but sits clearly below the exemplar descriptions in the rubric.

Suggestions

Add an explicit trigger clause, e.g. "Use when generating images from text prompts, performing inpainting or image-to-image transformations, or when the user mentions Stable Diffusion, diffusion models, or image generation."

Include natural trigger terms and synonyms users would say — "image generation", "Stable Diffusion", "diffusers", "outpainting" — to improve both completeness and trigger coverage.

Mention the library (HuggingFace Diffusers) to sharpen distinctiveness against other image-generation skills such as DALL-E or Flux.

Dimension	Reasoning	Score
Specificity	The description lists three concrete actions ("Text-to-image generation, inpainting, and img2img") comparable to the anchor example 'Extracts text from PDF files, fills forms, converts pages to images', though it omits library identity and body-level features like ControlNet, LoRA, and outpainting, which are more than trivial gaps but still minor relative to the core actions named.	4 / 5
Completeness	The 'what' is clear (three concrete capabilities), but there is no 'Use when...' clause or equivalent explicit trigger guidance, which per the judging guidelines caps completeness at 3.	3 / 5
Trigger Term Quality	"Text-to-image generation", "inpainting", and "img2img" are terms users naturally say, but the description misses common variations and synonyms such as "image generation", "Stable Diffusion", "outpainting", or "ControlNet", matching the anchor 'Some relevant keywords but missing common variations or synonyms'.	3 / 5
Distinctiveness Conflict Risk	The description is domain-specific to image generation but never names Stable Diffusion or the diffusers library, so it could overlap with other image-generation skills (e.g., DALL-E or Flux skills), fitting 'Somewhat specific but could still overlap with similar skills'.	3 / 5
	Total	13 / 20 Passed

Validation

75%

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 — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (524 lines); consider splitting into references/ and linking	Warning
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	12 / 16 Passed

Repository: NousResearch/hermes-agent
Path: optional-skills/mlops/stable-diffusion/SKILL.md
Commit: 68fa7e9

Reviewed: 3 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.