albumentations-team

Run performance benchmarks for transform changes. Use when the user asks to benchmark, measure performance, compare speed, or when changes affect apply methods, functional layer, parameter generation, or core pipeline code.

Policy for AlbumentationsX transforms that combine multiple images or objects. Use when implementing, reviewing, or using Mosaic, CopyAndPaste, OverlayElements, HistogramMatching, PixelDistributionAdaptation, or other mixing transforms.

Run the full shared Codex review checklist against a transform. Use when the user asks to review, audit, or check a transform for correctness, performance, or API consistency.

After completing code changes, runs tests and pre-commit, then iteratively fixes failures until all pass. Use when finishing a coding task, validating changes, or when the user asks to run tests or fix errors.

Full checklist for adding a new transform to AlbumentationsX. Use when the user asks to add, implement, or create a new transform/augmentation.

Quality bar for docstrings in albumentations. Use when writing or updating docstrings in albumentations/, especially for transforms and public APIs.

Systematic performance audit for AlbumentationsX runtime code. Use whenever implementing, reviewing, profiling, or optimizing transforms, functional kernels, apply methods, random generation, reductions, label maps, dtype conversions, batch paths,…

Use the repo `_internal/` directory for anything that must not be committed — scratch files, temporary outputs, local demos, Codex artifacts, or one-off scripts. Use when creating temp files, debug dumps, or local-only tooling during a task.

Validates whether benchmark artifacts cover the paper's required RGB micro and RGB DataLoader sections. Use when checking missing RGB runs, deciding what to run next, validating gcp_runs/output folders, or preparing paper tables.

Automates running image/video augmentation benchmarks for single or multiple libraries, validates outputs, generates comparison reports, and updates documentation. Use when running benchmarks, comparing library performance, or when the user mentions…

Updates benchmark documentation with latest results including README tables, speedup plots, and library metadata. Use when updating documentation, generating comparison tables, or when the user mentions update_docs.sh or documentation generation.

Triage detached GCP benchmark runs, DONE/FAILED sentinels, VM cleanup, vm.log, gcp_last_run.json, and partial result downloads. Use when GCP benchmark logs mention DONE, FAILED, exit_code.txt, VM disappeared, STOPPING, gcloud machine type errors, or missing…

Guides adding support for new image/video augmentation libraries to the benchmark suite. Use when integrating a new library, adding library support, or when the user mentions adding a new augmentation library to test.

Executes the paper benchmark plan for RGB, multichannel, DataLoader, and video benchmarks. Use when the user mentions the paper benchmark, deadline plan, machine matrix, RGB micro, multichannel, DataLoader, video GPU, c4/c4d/g2 machines, or what to run next.

Analyzes benchmark results to identify slow transforms, warmup issues, and performance regressions. Compares speedups across libraries and generates optimization recommendations. Use when analyzing performance, investigating slow benchmarks, or comparing…

Validates custom transform specification files for benchmark runner, checking LIBRARY, __call__, and TRANSFORMS definitions. Use when creating custom transforms, validating transform specs, or when the user mentions transform validation or spec files.

Albucore image processing conventions - shapes (H,W,C), dtypes (uint8/float32), benchmark-driven backend routing (OpenCV, NumPy, Torch CPU, LUT, NumKong), tests, and lockfile discipline. Use when implementing or modifying albucore modules, writing tests, or…

Running Albucore micro-benchmarks under benchmarks/, synthetic router timings, and comparing PyPI releases with uv --no-project. Use when adding benchmarks, comparing performance across versions, or documenting benchmark workflow.

Systematic performance audit for Albucore runtime code. Use whenever implementing, reviewing, profiling, or optimizing atomic image operations, backend routing, reductions, label maps, LUTs, random generation, dtype conversions, allocation-heavy paths, batch…

Optimize or review eager CPU-only Albucore PyTorch runtime paths with benchmark-backed decisions. Use when adding or changing Torch CPU kernels, Tensor/NumPy bridges, Torch backend routing, tensor layouts, allocations, threading, profiling, memory-format…

Albucore star-exported API (__all__), routers vs albucore.functions shims, and dependents such as Albumentations. Use when changing exports, documenting API, or deciding what belongs in package __all__.

Ensures Jupyter notebooks have required metadata and nbformat. Use when adding, creating, or modifying .ipynb files, or when the user asks about notebook structure or validation.