mflux-model-porting

Port ML models into mflux/MLX with correctness-first validation, then refactor toward mflux style.

Quality

67%

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 ./.cursor/skills/mflux-model-porting/SKILL.md

Quality

Content

81%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 dense, actionable, well-sequenced porting playbook with strong validation checkpoints and a clear integration checklist. It is efficient and concrete, with only minor room to tighten verbose sections and split off large reference blocks.

Dimension	Reasoning	Score
Conciseness	Efficient, domain-specific instruction with no padding of concepts Claude already knows; a few dense asides (e.g., the img2img paragraph) could be trimmed, keeping it just below a 5.	4 / 5
Actionability	Concrete commands (MFLUX_PRESERVE_TEST_OUTPUT=1 uv run, just lint, du -sh), specific paths, class names, and PR references make it mostly executable; minor gaps remain in a few conceptual steps like 'review recent ports'.	4 / 5
Workflow Clarity	A clear 7-step sequence with explicit validation checkpoints (deterministic tests, commit milestones, re-run tests during refactor, lint/test-fast before merge) plus a tick-list checklist and feedback loops.	5 / 5
Progressive Disclosure	Well-organized into clear sections (Goal, Principles, Workflow, Package layout, Integration checklist, Deliverables) with no nested references; no bundle files exist, and a couple of large blocks (package tree, integration tables) could plausibly live in separate files.	4 / 5
	Total	17 / 20 Passed

Description

53%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 clearly states what the skill does but omits any 'when to use it' trigger guidance, capping completeness. It is reasonably specific and distinct but would benefit from explicit trigger phrases.

Suggestions

Add a 'Use when ...' clause naming natural triggers (e.g., 'Use when porting a diffusers/HF model into mflux or when the user asks to bring a new model into MLX').

Expand the action list beyond port+refactor to mention weight mapping, deterministic testing, and CLI/config wiring for more comprehensive coverage.

Include synonyms/file hints users might say ('mflux port', 'MLX model', 'diffusers to mflux') to improve trigger term quality.

Dimension	Reasoning	Score
Specificity	Names the domain (mflux/MLX) and two concrete actions ("correctness-first validation" then "refactor toward mflux style"), but coverage is not comprehensive — only 1-2 actions.	3 / 5
Completeness	Has a clear 'what' (port models into mflux/MLX then refactor) but no 'Use when...' clause or equivalent trigger guidance, which caps completeness at 3 per the guideline.	3 / 5
Trigger Term Quality	Relevant keywords ("Port ML models", "mflux", "MLX") are present but it lacks common variations/synonyms and a natural trigger phrase a user would say.	3 / 5
Distinctiveness Conflict Risk	The mflux/MLX niche is specific and distinct; minor overlap risk with general ML-porting skills, but the correctness-first/refactor framing narrows it.	4 / 5
	Total	13 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: filipstrand/mflux
Path: .cursor/skills/mflux-model-porting/SKILL.md
Commit: 1430b27

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.