[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6MAYJ99E23MC2HSFQ20WTCB":3,"skill-same:01M6MAYJ99E23MC2HSFQ20WTCB":98},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":53,"linkOnly":54,"markdownContent":55,"markdownBody":56,"frontMatter":57,"refName":64,"commitSha":65,"skillDirectory":66,"forksCount":67,"repositoryHtmlUrl":68,"repositoryHomepageUrl":69,"ownerHtmlUrl":70,"ownerBlogUrl":71,"ownerDescription":72,"updatedAt":73,"install":74,"resources":87},"01M6MAYJ99E23MC2HSFQ20WTCB","ray-data",null,"Scalable data processing for ML workloads. Streaming execution across CPU\u002FGPU, supports Parquet\u002FCSV\u002FJSON\u002Fimages. Integrates with Ray Train, PyTorch, TensorFlow. Scales from single machine to 100s of nodes. Use for batch inference, data preprocessing, multi-modal data loading, or distributed ETL pipelines.","hashgraph-online\u002Fawesome-codex-plugins","hashgraph-online","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F160371962?v=4","01M6CNB6NRJVNNGN14TR01Q6D3","HOL (Hashgraph Online)","01M6JT1HSKKG9V68KQAQGBA7XP",1575,"MIT","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fskills\u002Fray-data\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000033","数据分析师","PROFESSION","适用职业",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52],"Ray Data","data","Ray","images","for","loading","Ray Train","Batch Inference","ML Pipelines","Scalable","CPU","processing","preprocessing","etl pipelines","distributed computing","across","train","inference","pipelines","use","json","pandas","multimodal","streaming","batch","etl","pytorch","distributed","tensorflow","2026-11-10 10:12:30",false,"---\nname: ray-data\ndescription: Scalable data processing for ML workloads. Streaming execution across CPU\u002FGPU, supports Parquet\u002FCSV\u002FJSON\u002Fimages. Integrates with Ray Train, PyTorch, TensorFlow. Scales from single machine to 100s of nodes. Use for batch inference, data preprocessing, multi-modal data loading, or distributed ETL pipelines.\nlicense: MIT\nmetadata:\n  version: 1.0.0\n  category: multimodal-and-emerging\n  maintainer: Kalaris Labs\n  tags: Data Processing, Ray Data, Distributed Computing, ML Pipelines, Batch Inference, ETL, Scalable, Ray, PyTorch, TensorFlow\n  dependencies: ray[data], pyarrow, pandas\n---\n\n# Ray Data - Scalable ML Data Processing\n\nDistributed data processing library for ML and AI workloads.\n\n## When to use Ray Data\n\n**Use Ray Data when:**\n- Processing large datasets (>100GB) for ML training\n- Need distributed data preprocessing across cluster\n- Building batch inference pipelines\n- Loading multi-modal data (images, audio, video)\n- Scaling data processing from laptop to cluster\n\n**Key features**:\n- **Streaming execution**: Process data larger than memory\n- **GPU support**: Accelerate transforms with GPUs\n- **Framework integration**: PyTorch, TensorFlow, HuggingFace\n- **Multi-modal**: Images, Parquet, CSV, JSON, audio, video\n\n**Use alternatives instead**:\n- **Pandas**: Small data (\u003C1GB) on single machine\n- **Dask**: Tabular data, SQL-like operations\n- **Spark**: Enterprise ETL, SQL queries\n\n## Quick start\n\n### Installation\n\n```bash\npip install -U 'ray[data]'\n```\n\n### Load and transform data\n\n```python\nimport ray\n\n# Read Parquet files\nds = ray.data.read_parquet(\"s3:\u002F\u002Fbucket\u002Fdata\u002F*.parquet\")\n\n# Transform data (lazy execution)\nds = ds.map_batches(lambda batch: {\"processed\": batch[\"text\"].str.lower()})\n\n# Consume data\nfor batch in ds.iter_batches(batch_size=100):\n    print(batch)\n```\n\n### Integration with Ray Train\n\n```python\nimport ray\nfrom ray.train import ScalingConfig\nfrom ray.train.torch import TorchTrainer\n\n# Create dataset\ntrain_ds = ray.data.read_parquet(\"s3:\u002F\u002Fbucket\u002Ftrain\u002F*.parquet\")\n\ndef train_func(config):\n    # Access dataset in training\n    train_ds = ray.train.get_dataset_shard(\"train\")\n\n    for epoch in range(10):\n        for batch in train_ds.iter_batches(batch_size=32):\n            # Train on batch\n            pass\n\n# Train with Ray\ntrainer = TorchTrainer(\n    train_func,\n    datasets={\"train\": train_ds},\n    scaling_config=ScalingConfig(num_workers=4, use_gpu=True)\n)\ntrainer.fit()\n```\n\n## Reading data\n\n### From cloud storage\n\n```python\nimport ray\n\n# Parquet (recommended for ML)\nds = ray.data.read_parquet(\"s3:\u002F\u002Fbucket\u002Fdata\u002F*.parquet\")\n\n# CSV\nds = ray.data.read_csv(\"s3:\u002F\u002Fbucket\u002Fdata\u002F*.csv\")\n\n# JSON\nds = ray.data.read_json(\"gs:\u002F\u002Fbucket\u002Fdata\u002F*.json\")\n\n# Images\nds = ray.data.read_images(\"s3:\u002F\u002Fbucket\u002Fimages\u002F\")\n```\n\n### From Python objects\n\n```python\n# From list\nds = ray.data.from_items([{\"id\": i, \"value\": i * 2} for i in range(1000)])\n\n# From range\nds = ray.data.range(1000000)  # Synthetic data\n\n# From pandas\nimport pandas as pd\ndf = pd.DataFrame({\"col1\": [1, 2, 3], \"col2\": [4, 5, 6]})\nds = ray.data.from_pandas(df)\n```\n\n## Transformations\n\n### Map batches (vectorized)\n\n```python\n# Batch transformation (fast)\ndef process_batch(batch):\n    batch[\"doubled\"] = batch[\"value\"] * 2\n    return batch\n\nds = ds.map_batches(process_batch, batch_size=1000)\n```\n\n### Row transformations\n\n```python\n# Row-by-row (slower)\ndef process_row(row):\n    row[\"squared\"] = row[\"value\"] ** 2\n    return row\n\nds = ds.map(process_row)\n```\n\n### Filter\n\n```python\n# Filter rows\nds = ds.filter(lambda row: row[\"value\"] > 100)\n```\n\n### Group by and aggregate\n\n```python\n# Group by column\nds = ds.groupby(\"category\").count()\n\n# Custom aggregation\nds = ds.groupby(\"category\").map_groups(lambda group: {\"sum\": group[\"value\"].sum()})\n```\n\n## GPU-accelerated transforms\n\n```python\n# Use GPU for preprocessing\ndef preprocess_images_gpu(batch):\n    import torch\n    images = torch.tensor(batch[\"image\"]).cuda()\n    # GPU preprocessing\n    processed = images * 255\n    return {\"processed\": processed.cpu().numpy()}\n\nds = ds.map_batches(\n    preprocess_images_gpu,\n    batch_size=64,\n    num_gpus=1  # Request GPU\n)\n```\n\n## Writing data\n\n```python\n# Write to Parquet\nds.write_parquet(\"s3:\u002F\u002Fbucket\u002Foutput\u002F\")\n\n# Write to CSV\nds.write_csv(\"output\u002F\")\n\n# Write to JSON\nds.write_json(\"output\u002F\")\n```\n\n## Performance optimization\n\n### Repartition\n\n```python\n# Control parallelism\nds = ds.repartition(100)  # 100 blocks for 100-core cluster\n```\n\n### Batch size tuning\n\n```python\n# Larger batches = faster vectorized ops\nds.map_batches(process_fn, batch_size=10000)  # vs batch_size=100\n```\n\n### Streaming execution\n\n```python\n# Process data larger than memory\nds = ray.data.read_parquet(\"s3:\u002F\u002Fhuge-dataset\u002F\")\nfor batch in ds.iter_batches(batch_size=1000):\n    process(batch)  # Streamed, not loaded to memory\n```\n\n## Common patterns\n\n### Batch inference\n\n```python\nimport ray\n\n# Load model\ndef load_model():\n    # Load once per worker\n    return MyModel()\n\n# Inference function\nclass BatchInference:\n    def __init__(self):\n        self.model = load_model()\n\n    def __call__(self, batch):\n        predictions = self.model(batch[\"input\"])\n        return {\"prediction\": predictions}\n\n# Run distributed inference\nds = ray.data.read_parquet(\"s3:\u002F\u002Fdata\u002F\")\npredictions = ds.map_batches(BatchInference, batch_size=32, num_gpus=1)\npredictions.write_parquet(\"s3:\u002F\u002Foutput\u002F\")\n```\n\n### Data preprocessing pipeline\n\n```python\n# Multi-step pipeline\nds = (\n    ray.data.read_parquet(\"s3:\u002F\u002Fraw\u002F\")\n    .map_batches(clean_data)\n    .map_batches(tokenize)\n    .map_batches(augment)\n    .write_parquet(\"s3:\u002F\u002Fprocessed\u002F\")\n)\n```\n\n## Integration with ML frameworks\n\n### PyTorch\n\n```python\n# Convert to PyTorch\ntorch_ds = ds.to_torch(label_column=\"label\", batch_size=32)\n\nfor batch in torch_ds:\n    # batch is dict with tensors\n    inputs, labels = batch[\"features\"], batch[\"label\"]\n```\n\n### TensorFlow\n\n```python\n# Convert to TensorFlow\ntf_ds = ds.to_tf(feature_columns=[\"image\"], label_column=\"label\", batch_size=32)\n\nfor features, labels in tf_ds:\n    # Train model\n    pass\n```\n\n## Supported data formats\n\n| Format | Read | Write | Use Case |\n|--------|------|-------|----------|\n| Parquet | ✅ | ✅ | ML data (recommended) |\n| CSV | ✅ | ✅ | Tabular data |\n| JSON | ✅ | ✅ | Semi-structured |\n| Images | ✅ | ❌ | Computer vision |\n| NumPy | ✅ | ✅ | Arrays |\n| Pandas | ✅ | ❌ | DataFrames |\n\n## Performance benchmarks\n\n**Scaling** (processing 100GB data):\n- 1 node (16 cores): ~30 minutes\n- 4 nodes (64 cores): ~8 minutes\n- 16 nodes (256 cores): ~2 minutes\n\n**GPU acceleration** (image preprocessing):\n- CPU only: 1,000 images\u002Fsec\n- 1 GPU: 5,000 images\u002Fsec\n- 4 GPUs: 18,000 images\u002Fsec\n\n## Use cases\n\n**Production deployments**:\n- **Pinterest**: Last-mile data processing for model training\n- **ByteDance**: Scaling offline inference with multi-modal LLMs\n- **Spotify**: ML platform for batch inference\n\n## References\n\n- **[Transformations Guide](references\u002Ftransformations.md)** - Map, filter, groupby operations\n- **[Integration Guide](references\u002Fintegration.md)** - Ray Train, PyTorch, TensorFlow\n\n## Resources\n\n- **Docs**: https:\u002F\u002Fdocs.ray.io\u002Fen\u002Flatest\u002Fdata\u002Fdata.html\n- **GitHub**: https:\u002F\u002Fgithub.com\u002Fray-project\u002Fray ⭐ 36,000+\n- **Version**: Ray 2.40.0+\n- **Examples**: https:\u002F\u002Fdocs.ray.io\u002Fen\u002Flatest\u002Fdata\u002Fexamples\u002Foverview.html\n\n## Agent operating procedure\n\n1. **Check the environment.** Confirm GPU memory, model checkpoint and licence, framework versions and input formats.\n2. **Pin down the inputs.** Confirm formats, identifiers and parameters from the data or the user. Ask rather than guess any value that changes the result.\n3. **Run a small version first.** Run inference on a single sample at low resolution or short length first.\n4. **Execute the full task** using the instructions and references above.\n5. **Validate the result.** Inspect outputs qualitatively and with task metrics; check licence restrictions for generated content.\n6. **Report.** State what was run (versions, commands, parameters), what was checked, and what is still uncertain.\n\n| If this happens | Do this |\n|---|---|\n| Out-of-memory on large inputs | Reduce resolution or sequence length, use half precision, or tile the input. |\n| A function, flag or endpoint in these instructions is missing in the installed version | Check the installed version's own documentation (`help()`, `--help`, official docs), adapt, and tell the user. Never invent an API. |\n| A required input, identifier or parameter is ambiguous | Ask the user, or state the assumption explicitly before running. |\n\n**Integrity rules**\n\n- Never fabricate results, parameters, identifiers, citations or statistics. If something cannot be run or verified, say so plainly.\n- Check model and dataset licences before commercial or published use.\n- Treat version-specific details here as possibly outdated: confirm them against the official documentation for the installed version.\n- Ask before actions that cost money, consume shared GPUs or cloud quota, touch personal or patient data, or cannot be undone.\n\n## Related skills\n\n- `ray-train`: Distributed training orchestration across clusters.\n- `polars`: High-performance DataFrame library for Python ETL, analytics, and pandas migration.\n- `dask`: Distributed computing for larger-than-RAM pandas\u002FNumPy workflows.\n","# Ray Data - Scalable ML Data Processing\n\nDistributed data processing library for ML and AI workloads.\n\n## When to use Ray Data\n\n**Use Ray Data when:**\n- Processing large datasets (>100GB) for ML training\n- Need distributed data preprocessing across cluster\n- Building batch inference pipelines\n- Loading multi-modal data (images, audio, video)\n- Scaling data processing from laptop to cluster\n\n**Key features**:\n- **Streaming execution**: Process data larger than memory\n- **GPU support**: Accelerate transforms with GPUs\n- **Framework integration**: PyTorch, TensorFlow, HuggingFace\n- **Multi-modal**: Images, Parquet, CSV, JSON, audio, video\n\n**Use alternatives instead**:\n- **Pandas**: Small data (\u003C1GB) on single machine\n- **Dask**: Tabular data, SQL-like operations\n- **Spark**: Enterprise ETL, SQL queries\n\n## Quick start\n\n### Installation\n\n```bash\npip install -U 'ray[data]'\n```\n\n### Load and transform data\n\n```python\nimport ray\n\n# Read Parquet files\nds = ray.data.read_parquet(\"s3:\u002F\u002Fbucket\u002Fdata\u002F*.parquet\")\n\n# Transform data (lazy execution)\nds = ds.map_batches(lambda batch: {\"processed\": batch[\"text\"].str.lower()})\n\n# Consume data\nfor batch in ds.iter_batches(batch_size=100):\n    print(batch)\n```\n\n### Integration with Ray Train\n\n```python\nimport ray\nfrom ray.train import ScalingConfig\nfrom ray.train.torch import TorchTrainer\n\n# Create dataset\ntrain_ds = ray.data.read_parquet(\"s3:\u002F\u002Fbucket\u002Ftrain\u002F*.parquet\")\n\ndef train_func(config):\n    # Access dataset in training\n    train_ds = ray.train.get_dataset_shard(\"train\")\n\n    for epoch in range(10):\n        for batch in train_ds.iter_batches(batch_size=32):\n            # Train on batch\n            pass\n\n# Train with Ray\ntrainer = TorchTrainer(\n    train_func,\n    datasets={\"train\": train_ds},\n    scaling_config=ScalingConfig(num_workers=4, use_gpu=True)\n)\ntrainer.fit()\n```\n\n## Reading data\n\n### From cloud storage\n\n```python\nimport ray\n\n# Parquet (recommended for ML)\nds = ray.data.read_parquet(\"s3:\u002F\u002Fbucket\u002Fdata\u002F*.parquet\")\n\n# CSV\nds = ray.data.read_csv(\"s3:\u002F\u002Fbucket\u002Fdata\u002F*.csv\")\n\n# JSON\nds = ray.data.read_json(\"gs:\u002F\u002Fbucket\u002Fdata\u002F*.json\")\n\n# Images\nds = ray.data.read_images(\"s3:\u002F\u002Fbucket\u002Fimages\u002F\")\n```\n\n### From Python objects\n\n```python\n# From list\nds = ray.data.from_items([{\"id\": i, \"value\": i * 2} for i in range(1000)])\n\n# From range\nds = ray.data.range(1000000)  # Synthetic data\n\n# From pandas\nimport pandas as pd\ndf = pd.DataFrame({\"col1\": [1, 2, 3], \"col2\": [4, 5, 6]})\nds = ray.data.from_pandas(df)\n```\n\n## Transformations\n\n### Map batches (vectorized)\n\n```python\n# Batch transformation (fast)\ndef process_batch(batch):\n    batch[\"doubled\"] = batch[\"value\"] * 2\n    return batch\n\nds = ds.map_batches(process_batch, batch_size=1000)\n```\n\n### Row transformations\n\n```python\n# Row-by-row (slower)\ndef process_row(row):\n    row[\"squared\"] = row[\"value\"] ** 2\n    return row\n\nds = ds.map(process_row)\n```\n\n### Filter\n\n```python\n# Filter rows\nds = ds.filter(lambda row: row[\"value\"] > 100)\n```\n\n### Group by and aggregate\n\n```python\n# Group by column\nds = ds.groupby(\"category\").count()\n\n# Custom aggregation\nds = ds.groupby(\"category\").map_groups(lambda group: {\"sum\": group[\"value\"].sum()})\n```\n\n## GPU-accelerated transforms\n\n```python\n# Use GPU for preprocessing\ndef preprocess_images_gpu(batch):\n    import torch\n    images = torch.tensor(batch[\"image\"]).cuda()\n    # GPU preprocessing\n    processed = images * 255\n    return {\"processed\": processed.cpu().numpy()}\n\nds = ds.map_batches(\n    preprocess_images_gpu,\n    batch_size=64,\n    num_gpus=1  # Request GPU\n)\n```\n\n## Writing data\n\n```python\n# Write to Parquet\nds.write_parquet(\"s3:\u002F\u002Fbucket\u002Foutput\u002F\")\n\n# Write to CSV\nds.write_csv(\"output\u002F\")\n\n# Write to JSON\nds.write_json(\"output\u002F\")\n```\n\n## Performance optimization\n\n### Repartition\n\n```python\n# Control parallelism\nds = ds.repartition(100)  # 100 blocks for 100-core cluster\n```\n\n### Batch size tuning\n\n```python\n# Larger batches = faster vectorized ops\nds.map_batches(process_fn, batch_size=10000)  # vs batch_size=100\n```\n\n### Streaming execution\n\n```python\n# Process data larger than memory\nds = ray.data.read_parquet(\"s3:\u002F\u002Fhuge-dataset\u002F\")\nfor batch in ds.iter_batches(batch_size=1000):\n    process(batch)  # Streamed, not loaded to memory\n```\n\n## Common patterns\n\n### Batch inference\n\n```python\nimport ray\n\n# Load model\ndef load_model():\n    # Load once per worker\n    return MyModel()\n\n# Inference function\nclass BatchInference:\n    def __init__(self):\n        self.model = load_model()\n\n    def __call__(self, batch):\n        predictions = self.model(batch[\"input\"])\n        return {\"prediction\": predictions}\n\n# Run distributed inference\nds = ray.data.read_parquet(\"s3:\u002F\u002Fdata\u002F\")\npredictions = ds.map_batches(BatchInference, batch_size=32, num_gpus=1)\npredictions.write_parquet(\"s3:\u002F\u002Foutput\u002F\")\n```\n\n### Data preprocessing pipeline\n\n```python\n# Multi-step pipeline\nds = (\n    ray.data.read_parquet(\"s3:\u002F\u002Fraw\u002F\")\n    .map_batches(clean_data)\n    .map_batches(tokenize)\n    .map_batches(augment)\n    .write_parquet(\"s3:\u002F\u002Fprocessed\u002F\")\n)\n```\n\n## Integration with ML frameworks\n\n### PyTorch\n\n```python\n# Convert to PyTorch\ntorch_ds = ds.to_torch(label_column=\"label\", batch_size=32)\n\nfor batch in torch_ds:\n    # batch is dict with tensors\n    inputs, labels = batch[\"features\"], batch[\"label\"]\n```\n\n### TensorFlow\n\n```python\n# Convert to TensorFlow\ntf_ds = ds.to_tf(feature_columns=[\"image\"], label_column=\"label\", batch_size=32)\n\nfor features, labels in tf_ds:\n    # Train model\n    pass\n```\n\n## Supported data formats\n\n| Format | Read | Write | Use Case |\n|--------|------|-------|----------|\n| Parquet | ✅ | ✅ | ML data (recommended) |\n| CSV | ✅ | ✅ | Tabular data |\n| JSON | ✅ | ✅ | Semi-structured |\n| Images | ✅ | ❌ | Computer vision |\n| NumPy | ✅ | ✅ | Arrays |\n| Pandas | ✅ | ❌ | DataFrames |\n\n## Performance benchmarks\n\n**Scaling** (processing 100GB data):\n- 1 node (16 cores): ~30 minutes\n- 4 nodes (64 cores): ~8 minutes\n- 16 nodes (256 cores): ~2 minutes\n\n**GPU acceleration** (image preprocessing):\n- CPU only: 1,000 images\u002Fsec\n- 1 GPU: 5,000 images\u002Fsec\n- 4 GPUs: 18,000 images\u002Fsec\n\n## Use cases\n\n**Production deployments**:\n- **Pinterest**: Last-mile data processing for model training\n- **ByteDance**: Scaling offline inference with multi-modal LLMs\n- **Spotify**: ML platform for batch inference\n\n## References\n\n- **[Transformations Guide](references\u002Ftransformations.md)** - Map, filter, groupby operations\n- **[Integration Guide](references\u002Fintegration.md)** - Ray Train, PyTorch, TensorFlow\n\n## Resources\n\n- **Docs**: https:\u002F\u002Fdocs.ray.io\u002Fen\u002Flatest\u002Fdata\u002Fdata.html\n- **GitHub**: https:\u002F\u002Fgithub.com\u002Fray-project\u002Fray ⭐ 36,000+\n- **Version**: Ray 2.40.0+\n- **Examples**: https:\u002F\u002Fdocs.ray.io\u002Fen\u002Flatest\u002Fdata\u002Fexamples\u002Foverview.html\n\n## Agent operating procedure\n\n1. **Check the environment.** Confirm GPU memory, model checkpoint and licence, framework versions and input formats.\n2. **Pin down the inputs.** Confirm formats, identifiers and parameters from the data or the user. Ask rather than guess any value that changes the result.\n3. **Run a small version first.** Run inference on a single sample at low resolution or short length first.\n4. **Execute the full task** using the instructions and references above.\n5. **Validate the result.** Inspect outputs qualitatively and with task metrics; check licence restrictions for generated content.\n6. **Report.** State what was run (versions, commands, parameters), what was checked, and what is still uncertain.\n\n| If this happens | Do this |\n|---|---|\n| Out-of-memory on large inputs | Reduce resolution or sequence length, use half precision, or tile the input. |\n| A function, flag or endpoint in these instructions is missing in the installed version | Check the installed version's own documentation (`help()`, `--help`, official docs), adapt, and tell the user. Never invent an API. |\n| A required input, identifier or parameter is ambiguous | Ask the user, or state the assumption explicitly before running. |\n\n**Integrity rules**\n\n- Never fabricate results, parameters, identifiers, citations or statistics. If something cannot be run or verified, say so plainly.\n- Check model and dataset licences before commercial or published use.\n- Treat version-specific details here as possibly outdated: confirm them against the official documentation for the installed version.\n- Ask before actions that cost money, consume shared GPUs or cloud quota, touch personal or patient data, or cannot be undone.\n\n## Related skills\n\n- `ray-train`: Distributed training orchestration across clusters.\n- `polars`: High-performance DataFrame library for Python ETL, analytics, and pandas migration.\n- `dask`: Distributed computing for larger-than-RAM pandas\u002FNumPy workflows.\n",{"name":5,"license":15,"metadata":58,"description":7},{"tags":59,"version":60,"category":61,"maintainer":62,"dependencies":63},"Data Processing, Ray Data, Distributed Computing, ML Pipelines, Batch Inference, ETL, Scalable, Ray, PyTorch, TensorFlow","1.0.0","multimodal-and-emerging","Kalaris Labs","ray[data], pyarrow, pandas","main","2b63f856ce19c2a797929cacabd9e0fc936812fa","plugins\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fskills\u002Fray-data",491,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins","https:\u002F\u002Fhol.org\u002Fplugins\u002Fbest-codex-plugins","https:\u002F\u002Fgithub.com\u002Fhashgraph-online","https:\u002F\u002Fhol.org","Open-source standards and developer tools for interoperable AI agents: discovery, identity, communication, payments, privacy and security.","2026-11-14 14:06:13",{"source":8,"skillName":5,"commands":75,"prompt":85,"zipAvailable":86},[76,79,82],{"tool":77,"command":78},"npx","npx skills add hashgraph-online\u002Fawesome-codex-plugins --skill ray-data",{"tool":80,"command":81},"bunx","bunx skills add hashgraph-online\u002Fawesome-codex-plugins --skill ray-data",{"tool":83,"command":84},"pnpm","pnpm dlx skills add hashgraph-online\u002Fawesome-codex-plugins --skill ray-data","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fskills\u002Fray-data\u002FSKILL.md",true,[88,94],{"path":89,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":92,"sourceHtmlUrl":93},"references\u002Fintegration.md","REFERENCE","参考文档",1851,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fskills\u002Fray-data\u002Freferences\u002Fintegration.md",{"path":95,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":96,"sourceHtmlUrl":97},"references\u002Ftransformations.md",1664,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fskills\u002Fray-data\u002Freferences\u002Ftransformations.md",[99,108,116,124,132,140],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":103,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":107},"01M7DV332YQMGD8WTRJ9DZQXTB","worktable-review-with-annotations","Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-review-with-annotations\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-review-with-annotations\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":109,"name":110,"nameZh":6,"description":111,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":112,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":107},"01M7DV2YGHCZVBEB1AHYB6CGHE","worktable-find-and-synthesize-context","Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-find-and-synthesize-context\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-find-and-synthesize-context\u002FSKILL.md",[],[],{"skillNo":117,"name":118,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":120,"sourceHtmlUrl":121,"categories":122,"keywords":123,"publishedAt":107},"01M7DV2TQN1H11ME2SAF9MXN7V","worktable-create-or-update-docs","Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-update-docs\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-update-docs\u002FSKILL.md",[],[],{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":128,"sourceHtmlUrl":129,"categories":130,"keywords":131,"publishedAt":107},"01M7DV2NE452NEA8TXATKSNAJH","worktable-create-or-manage-records","Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-records\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-records\u002FSKILL.md",[],[],{"skillNo":133,"name":134,"nameZh":6,"description":135,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":136,"sourceHtmlUrl":137,"categories":138,"keywords":139,"publishedAt":107},"01M7DV2HENRBR2KSCVM064WH0Y","worktable-create-or-manage-interactive-html","Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-interactive-html\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-interactive-html\u002FSKILL.md",[],[],{"skillNo":141,"name":142,"nameZh":6,"description":143,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":144,"sourceHtmlUrl":145,"categories":146,"keywords":147,"publishedAt":107},"01M7DV2CADR5R7RQFFYPNAGR9Y","worktable-collaborate-threads","Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-collaborate-threads\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-collaborate-threads\u002FSKILL.md",[],[]]