mflux-community
Official@mflux-community
Apple MLX native implementations of state-of-the-art generative image & video models
Agent Skills by mflux-community
Showing 10 vetted skills indexed across 1 GitHub repositories.
mflux-communitymflux-cli
Navigate MFLUX CLI commands and inventory supported image generation features.
mflux-communitymflux-testing
Runs pytest suites in mflux and manages golden image test outputs safely.
mflux-communitymflux-model-porting
Port ML models from diffusers into mflux/MLX with correctness-first deterministic validation.
mflux-communitymflux-model-tiny-test
Writes hermetic tiny model-saving tests for mflux models at toy dimensions.
mflux-communitymflux-debugging
Diagnose MLX port divergences by comparing exported tensors against PyTorch reference runs.
mflux-communitymflux-dev-env
Configures the mflux development environment using uv, justfile recipes, and ty type checking.
mflux-communitymflux-release
Prepares mflux release branches with version bumps and uv lockfile updates.
mflux-communitymflux-manual-testing
Validates mflux CLI changes by running commands and reviewing generated images and artifacts.
mflux-communitymflux-pr
Guides pull request creation in the mflux repository with lint, test, and diff audit checks.
mflux-communityremediating-dependabot
Remediates GitHub Dependabot alerts by updating pyproject.toml and uv.lock dependencies.
Frequently Asked Questions About mflux-community
FAQPage SchemaWhat tasks can I perform using mflux-community's skills?▼
You can navigate the MFLUX CLI, port ML models into MLX, write hermetic tiny model-saving tests, run fast/slow/full test suites with golden image diffs, debug ports against PyTorch references, set up the dev environment, prepare releases, open clean PRs, and remediate Dependabot alerts.
Who are mflux-community's skills designed for?▼
They target ML engineers and contributors working on Apple MLX-native generative image and video models, specifically developers porting models from PyTorch/diffusers, maintaining the mflux codebase, validating output images, and managing releases and dependency security.
What is the practical development workflow for mflux?▼
Set up the arm64 dev environment with uv and justfile recipes for lint, format, and test. Make changes, run fast or full tests preserving image outputs, manually validate changed CLI paths by reviewing output artifacts, then inspect the diff, commit, push, and open a PR.
What prerequisites and dependencies does mflux development require?▼
An arm64 Apple Silicon environment is expected, with uv for dependency management and justfile recipes for linting, formatting, and testing. Dependencies are declared in pyproject.toml and locked in uv.lock, with Dependabot alerts remediated via clamping or upgrading those files.
How does mflux handle testing and release management?▼
Testing offers fast, slow, and full suites with golden image diff handling plus hermetic tiny save/load tests at toy dimensions. Releases involve a version bump and uv lock update, with release notes harvested from PR release-note blocks and published in GitHub Releases.