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

77%Weight 40%Scale 1-3

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

A highly actionable, well-sequenced porting workflow packed with mflux-specific specifics and validation checkpoints, held back only by monolithic inlining of large tables/layout that could be split into reference files.

Suggestions

Move the 'Integration surfaces checklist' tables and/or 'Package layout' into separate reference files (e.g. INTEGRATION.md, LAYOUT.md) and link to them from SKILL.md to improve progressive disclosure.

Tighten the longest tables to the rows most likely to be missed, pruning redundant detail to reduce token load.

Consider extracting the per-surface PR reference list into a reference doc so the main body stays a lean overview.

Dimension	Reasoning	Score
Conciseness	The body is information-dense with mflux-specific knowledge Claude lacks (file paths, PR numbers, gotchas) and avoids generic concept explanations, but at ~190 lines with several long tables it is longer than strictly necessary and could be tightened.	2 / 3
Actionability	Gives concrete, executable guidance throughout — exact commands ('MFLUX_PRESERVE_TEST_OUTPUT=1 uv run', 'make lint', 'du -sh'), specific file paths, class names, and PR references — making the instruction actionable rather than abstract.	3 / 3
Workflow Clarity	A clearly sequenced 7-step workflow with explicit validation checkpoints ('deterministic validation', 're-run make lint / make test-fast', 'keep running the deterministic image test during refactors') and feedback loops for error recovery.	3 / 3
Progressive Disclosure	No bundle files exist and the entire ~190-line skill lives in one monolithic SKILL.md; sections are well organized, but the integration tables and package layout are inline content that arguably belongs in separate reference files.	2 / 3
	Total	10 / 12 Passed

Description

57%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A focused, third-person description with a clear niche, but it omits an explicit 'Use when' trigger clause and lists only two high-level actions, so it lands at the middle band on specificity, trigger quality, and completeness.

Suggestions

Add an explicit trigger clause, e.g. 'Use when porting or translating an ML model from diffusers/PyTorch into mflux (MLX)'.

Name a few more concrete actions to lift specificity (e.g. 'export reference latents, map weights, validate parity, wire CLI').

Include common user phrasings ('convert', 'translate', 'MLX port') alongside 'Port' to broaden trigger coverage.

Dimension	Reasoning	Score
Specificity	Names the domain (ML models, mflux/MLX) and two actions ('Port ... with correctness-first validation, then refactor'), but stops at two rather than listing multiple concrete actions.	2 / 3
Completeness	Answers 'what' clearly but has no 'Use when ...' clause or equivalent explicit trigger guidance, which caps completeness at 2 per the guidelines.	2 / 3
Trigger Term Quality	Includes relevant terms a user might say ('Port ML models', 'mflux/MLX', 'refactor') but leans technical and omits common phrasings like 'convert', 'translate', or 'MLX port'.	2 / 3
Distinctiveness Conflict Risk	The mflux/MLX niche is narrow and specific, making it unlikely to trigger for unrelated skills despite the missing explicit trigger.	3 / 3
	Total	9 / 12 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: 4ac641e

Reviewed: 8 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.