[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6NTZPMWC4DN3RFJQ5N8ESB3":3,"skill-same:01M6NTZPMWC4DN3RFJQ5N8ESB3":60},{"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":18,"publishedAt":33,"linkOnly":34,"markdownContent":35,"markdownBody":36,"frontMatter":37,"refName":39,"commitSha":40,"skillDirectory":41,"forksCount":42,"repositoryHtmlUrl":43,"repositoryHomepageUrl":6,"ownerHtmlUrl":44,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":45,"install":46,"resources":59},"01M6NTZPMWC4DN3RFJQ5N8ESB3","machine-learning-ops-ml-pipeline",null,"Design and implement a complete ML pipeline for: $ARGUMENTS","mattonweb\u002Fai-assistant-foundry","mattonweb","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F2309734?v=4","01M6NQYCY846TKYN2ZCRS71WQ7","Matt Johnson","01M6NQYD2NPAQEFVD8TW9W0SVX",1,"MIT","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fdata-ml\u002Fsickn33\u002Fmachine-learning-ops-ml-pipeline\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32],"machine learning","ml","pipeline","learning","ops","machine learning ops ml pipeline","ml pipeline","machine","design","implement","community","complete","for","unknown","2026-11-10 10:12:30",false,"---\nname: machine-learning-ops-ml-pipeline\ndescription: \"Design and implement a complete ML pipeline for: $ARGUMENTS\"\nrisk: unknown\nsource: community\ndate_added: \"2026-02-27\"\n---\n\n# Machine Learning Pipeline - Multi-Agent MLOps Orchestration\n\nDesign and implement a complete ML pipeline for: $ARGUMENTS\n\n## Use this skill when\n\n- Working on machine learning pipeline - multi-agent mlops orchestration tasks or workflows\n- Needing guidance, best practices, or checklists for machine learning pipeline - multi-agent mlops orchestration\n\n## Do not use this skill when\n\n- The task is unrelated to machine learning pipeline - multi-agent mlops orchestration\n- You need a different domain or tool outside this scope\n\n## Instructions\n\n- Clarify goals, constraints, and required inputs.\n- Apply relevant best practices and validate outcomes.\n- Provide actionable steps and verification.\n- If detailed examples are required, open `resources\u002Fimplementation-playbook.md`.\n\n## Thinking\n\nThis workflow orchestrates multiple specialized agents to build a production-ready ML pipeline following modern MLOps best practices. The approach emphasizes:\n\n- **Phase-based coordination**: Each phase builds upon previous outputs, with clear handoffs between agents\n- **Modern tooling integration**: MLflow\u002FW&B for experiments, Feast\u002FTecton for features, KServe\u002FSeldon for serving\n- **Production-first mindset**: Every component designed for scale, monitoring, and reliability\n- **Reproducibility**: Version control for data, models, and infrastructure\n- **Continuous improvement**: Automated retraining, A\u002FB testing, and drift detection\n\nThe multi-agent approach ensures each aspect is handled by domain experts:\n- Data engineers handle ingestion and quality\n- Data scientists design features and experiments\n- ML engineers implement training pipelines\n- MLOps engineers handle production deployment\n- Observability engineers ensure monitoring\n\n## Phase 1: Data & Requirements Analysis\n\n\u003CTask>\nsubagent_type: data-engineer\nprompt: |\n  Analyze and design data pipeline for ML system with requirements: $ARGUMENTS\n\n  Deliverables:\n  1. Data source audit and ingestion strategy:\n     - Source systems and connection patterns\n     - Schema validation using Pydantic\u002FGreat Expectations\n     - Data versioning with DVC or lakeFS\n     - Incremental loading and CDC strategies\n\n  2. Data quality framework:\n     - Profiling and statistics generation\n     - Anomaly detection rules\n     - Data lineage tracking\n     - Quality gates and SLAs\n\n  3. Storage architecture:\n     - Raw\u002Fprocessed\u002Ffeature layers\n     - Partitioning strategy\n     - Retention policies\n     - Cost optimization\n\n  Provide implementation code for critical components and integration patterns.\n\u003C\u002FTask>\n\n\u003CTask>\nsubagent_type: data-scientist\nprompt: |\n  Design feature engineering and model requirements for: $ARGUMENTS\n  Using data architecture from: {phase1.data-engineer.output}\n\n  Deliverables:\n  1. Feature engineering pipeline:\n     - Transformation specifications\n     - Feature store schema (Feast\u002FTecton)\n     - Statistical validation rules\n     - Handling strategies for missing data\u002Foutliers\n\n  2. Model requirements:\n     - Algorithm selection rationale\n     - Performance metrics and baselines\n     - Training data requirements\n     - Evaluation criteria and thresholds\n\n  3. Experiment design:\n     - Hypothesis and success metrics\n     - A\u002FB testing methodology\n     - Sample size calculations\n     - Bias detection approach\n\n  Include feature transformation code and statistical validation logic.\n\u003C\u002FTask>\n\n## Phase 2: Model Development & Training\n\n\u003CTask>\nsubagent_type: ml-engineer\nprompt: |\n  Implement training pipeline based on requirements: {phase1.data-scientist.output}\n  Using data pipeline: {phase1.data-engineer.output}\n\n  Build comprehensive training system:\n  1. Training pipeline implementation:\n     - Modular training code with clear interfaces\n     - Hyperparameter optimization (Optuna\u002FRay Tune)\n     - Distributed training support (Horovod\u002FPyTorch DDP)\n     - Cross-validation and ensemble strategies\n\n  2. Experiment tracking setup:\n     - MLflow\u002FWeights & Biases integration\n     - Metric logging and visualization\n     - Artifact management (models, plots, data samples)\n     - Experiment comparison and analysis tools\n\n  3. Model registry integration:\n     - Version control and tagging strategy\n     - Model metadata and lineage\n     - Promotion workflows (dev -> staging -> prod)\n     - Rollback procedures\n\n  Provide complete training code with configuration management.\n\u003C\u002FTask>\n\n\u003CTask>\nsubagent_type: python-pro\nprompt: |\n  Optimize and productionize ML code from: {phase2.ml-engineer.output}\n\n  Focus areas:\n  1. Code quality and structure:\n     - Refactor for production standards\n     - Add comprehensive error handling\n     - Implement proper logging with structured formats\n     - Create reusable components and utilities\n\n  2. Performance optimization:\n     - Profile and optimize bottlenecks\n     - Implement caching strategies\n     - Optimize data loading and preprocessing\n     - Memory management for large-scale training\n\n  3. Testing framework:\n     - Unit tests for data transformations\n     - Integration tests for pipeline components\n     - Model quality tests (invariance, directional)\n     - Performance regression tests\n\n  Deliver production-ready, maintainable code with full test coverage.\n\u003C\u002FTask>\n\n## Phase 3: Production Deployment & Serving\n\n\u003CTask>\nsubagent_type: mlops-engineer\nprompt: |\n  Design production deployment for models from: {phase2.ml-engineer.output}\n  With optimized code from: {phase2.python-pro.output}\n\n  Implementation requirements:\n  1. Model serving infrastructure:\n     - REST\u002FgRPC APIs with FastAPI\u002FTorchServe\n     - Batch prediction pipelines (Airflow\u002FKubeflow)\n     - Stream processing (Kafka\u002FKinesis integration)\n     - Model serving platforms (KServe\u002FSeldon Core)\n\n  2. Deployment strategies:\n     - Blue-green deployments for zero downtime\n     - Canary releases with traffic splitting\n     - Shadow deployments for validation\n     - A\u002FB testing infrastructure\n\n  3. CI\u002FCD pipeline:\n     - GitHub Actions\u002FGitLab CI workflows\n     - Automated testing gates\n     - Model validation before deployment\n     - ArgoCD for GitOps deployment\n\n  4. Infrastructure as Code:\n     - Terraform modules for cloud resources\n     - Helm charts for Kubernetes deployments\n     - Docker multi-stage builds for optimization\n     - Secret management with Vault\u002FSecrets Manager\n\n  Provide complete deployment configuration and automation scripts.\n\u003C\u002FTask>\n\n\u003CTask>\nsubagent_type: kubernetes-architect\nprompt: |\n  Design Kubernetes infrastructure for ML workloads from: {phase3.mlops-engineer.output}\n\n  Kubernetes-specific requirements:\n  1. Workload orchestration:\n     - Training job scheduling with Kubeflow\n     - GPU resource allocation and sharing\n     - Spot\u002Fpreemptible instance integration\n     - Priority classes and resource quotas\n\n  2. Serving infrastructure:\n     - HPA\u002FVPA for autoscaling\n     - KEDA for event-driven scaling\n     - Istio service mesh for traffic management\n     - Model caching and warm-up strategies\n\n  3. Storage and data access:\n     - PVC strategies for training data\n     - Model artifact storage with CSI drivers\n     - Distributed storage for feature stores\n     - Cache layers for inference optimization\n\n  Provide Kubernetes manifests and Helm charts for entire ML platform.\n\u003C\u002FTask>\n\n## Phase 4: Monitoring & Continuous Improvement\n\n\u003CTask>\nsubagent_type: observability-engineer\nprompt: |\n  Implement comprehensive monitoring for ML system deployed in: {phase3.mlops-engineer.output}\n  Using Kubernetes infrastructure: {phase3.kubernetes-architect.output}\n\n  Monitoring framework:\n  1. Model performance monitoring:\n     - Prediction accuracy tracking\n     - Latency and throughput metrics\n     - Feature importance shifts\n     - Business KPI correlation\n\n  2. Data and model drift detection:\n     - Statistical drift detection (KS test, PSI)\n     - Concept drift monitoring\n     - Feature distribution tracking\n     - Automated drift alerts and reports\n\n  3. System observability:\n     - Prometheus metrics for all components\n     - Grafana dashboards for visualization\n     - Distributed tracing with Jaeger\u002FZipkin\n     - Log aggregation with ELK\u002FLoki\n\n  4. Alerting and automation:\n     - PagerDuty\u002FOpsgenie integration\n     - Automated retraining triggers\n     - Performance degradation workflows\n     - Incident response runbooks\n\n  5. Cost tracking:\n     - Resource utilization metrics\n     - Cost allocation by model\u002Fexperiment\n     - Optimization recommendations\n     - Budget alerts and controls\n\n  Deliver monitoring configuration, dashboards, and alert rules.\n\u003C\u002FTask>\n\n## Configuration Options\n\n- **experiment_tracking**: mlflow | wandb | neptune | clearml\n- **feature_store**: feast | tecton | databricks | custom\n- **serving_platform**: kserve | seldon | torchserve | triton\n- **orchestration**: kubeflow | airflow | prefect | dagster\n- **cloud_provider**: aws | azure | gcp | multi-cloud\n- **deployment_mode**: realtime | batch | streaming | hybrid\n- **monitoring_stack**: prometheus | datadog | newrelic | custom\n\n## Success Criteria\n\n1. **Data Pipeline Success**:\n   - \u003C 0.1% data quality issues in production\n   - Automated data validation passing 99.9% of time\n   - Complete data lineage tracking\n   - Sub-second feature serving latency\n\n2. **Model Performance**:\n   - Meeting or exceeding baseline metrics\n   - \u003C 5% performance degradation before retraining\n   - Successful A\u002FB tests with statistical significance\n   - No undetected model drift > 24 hours\n\n3. **Operational Excellence**:\n   - 99.9% uptime for model serving\n   - \u003C 200ms p99 inference latency\n   - Automated rollback within 5 minutes\n   - Complete observability with \u003C 1 minute alert time\n\n4. **Development Velocity**:\n   - \u003C 1 hour from commit to production\n   - Parallel experiment execution\n   - Reproducible training runs\n   - Self-service model deployment\n\n5. **Cost Efficiency**:\n   - \u003C 20% infrastructure waste\n   - Optimized resource allocation\n   - Automatic scaling based on load\n   - Spot instance utilization > 60%\n\n## Final Deliverables\n\nUpon completion, the orchestrated pipeline will provide:\n- End-to-end ML pipeline with full automation\n- Comprehensive documentation and runbooks\n- Production-ready infrastructure as code\n- Complete monitoring and alerting system\n- CI\u002FCD pipelines for continuous improvement\n- Cost optimization and scaling strategies\n- Disaster recovery and rollback procedures\n","# Machine Learning Pipeline - Multi-Agent MLOps Orchestration\n\nDesign and implement a complete ML pipeline for: $ARGUMENTS\n\n## Use this skill when\n\n- Working on machine learning pipeline - multi-agent mlops orchestration tasks or workflows\n- Needing guidance, best practices, or checklists for machine learning pipeline - multi-agent mlops orchestration\n\n## Do not use this skill when\n\n- The task is unrelated to machine learning pipeline - multi-agent mlops orchestration\n- You need a different domain or tool outside this scope\n\n## Instructions\n\n- Clarify goals, constraints, and required inputs.\n- Apply relevant best practices and validate outcomes.\n- Provide actionable steps and verification.\n- If detailed examples are required, open `resources\u002Fimplementation-playbook.md`.\n\n## Thinking\n\nThis workflow orchestrates multiple specialized agents to build a production-ready ML pipeline following modern MLOps best practices. The approach emphasizes:\n\n- **Phase-based coordination**: Each phase builds upon previous outputs, with clear handoffs between agents\n- **Modern tooling integration**: MLflow\u002FW&B for experiments, Feast\u002FTecton for features, KServe\u002FSeldon for serving\n- **Production-first mindset**: Every component designed for scale, monitoring, and reliability\n- **Reproducibility**: Version control for data, models, and infrastructure\n- **Continuous improvement**: Automated retraining, A\u002FB testing, and drift detection\n\nThe multi-agent approach ensures each aspect is handled by domain experts:\n- Data engineers handle ingestion and quality\n- Data scientists design features and experiments\n- ML engineers implement training pipelines\n- MLOps engineers handle production deployment\n- Observability engineers ensure monitoring\n\n## Phase 1: Data & Requirements Analysis\n\n\u003CTask>\nsubagent_type: data-engineer\nprompt: |\n  Analyze and design data pipeline for ML system with requirements: $ARGUMENTS\n\n  Deliverables:\n  1. Data source audit and ingestion strategy:\n     - Source systems and connection patterns\n     - Schema validation using Pydantic\u002FGreat Expectations\n     - Data versioning with DVC or lakeFS\n     - Incremental loading and CDC strategies\n\n  2. Data quality framework:\n     - Profiling and statistics generation\n     - Anomaly detection rules\n     - Data lineage tracking\n     - Quality gates and SLAs\n\n  3. Storage architecture:\n     - Raw\u002Fprocessed\u002Ffeature layers\n     - Partitioning strategy\n     - Retention policies\n     - Cost optimization\n\n  Provide implementation code for critical components and integration patterns.\n\u003C\u002FTask>\n\n\u003CTask>\nsubagent_type: data-scientist\nprompt: |\n  Design feature engineering and model requirements for: $ARGUMENTS\n  Using data architecture from: {phase1.data-engineer.output}\n\n  Deliverables:\n  1. Feature engineering pipeline:\n     - Transformation specifications\n     - Feature store schema (Feast\u002FTecton)\n     - Statistical validation rules\n     - Handling strategies for missing data\u002Foutliers\n\n  2. Model requirements:\n     - Algorithm selection rationale\n     - Performance metrics and baselines\n     - Training data requirements\n     - Evaluation criteria and thresholds\n\n  3. Experiment design:\n     - Hypothesis and success metrics\n     - A\u002FB testing methodology\n     - Sample size calculations\n     - Bias detection approach\n\n  Include feature transformation code and statistical validation logic.\n\u003C\u002FTask>\n\n## Phase 2: Model Development & Training\n\n\u003CTask>\nsubagent_type: ml-engineer\nprompt: |\n  Implement training pipeline based on requirements: {phase1.data-scientist.output}\n  Using data pipeline: {phase1.data-engineer.output}\n\n  Build comprehensive training system:\n  1. Training pipeline implementation:\n     - Modular training code with clear interfaces\n     - Hyperparameter optimization (Optuna\u002FRay Tune)\n     - Distributed training support (Horovod\u002FPyTorch DDP)\n     - Cross-validation and ensemble strategies\n\n  2. Experiment tracking setup:\n     - MLflow\u002FWeights & Biases integration\n     - Metric logging and visualization\n     - Artifact management (models, plots, data samples)\n     - Experiment comparison and analysis tools\n\n  3. Model registry integration:\n     - Version control and tagging strategy\n     - Model metadata and lineage\n     - Promotion workflows (dev -> staging -> prod)\n     - Rollback procedures\n\n  Provide complete training code with configuration management.\n\u003C\u002FTask>\n\n\u003CTask>\nsubagent_type: python-pro\nprompt: |\n  Optimize and productionize ML code from: {phase2.ml-engineer.output}\n\n  Focus areas:\n  1. Code quality and structure:\n     - Refactor for production standards\n     - Add comprehensive error handling\n     - Implement proper logging with structured formats\n     - Create reusable components and utilities\n\n  2. Performance optimization:\n     - Profile and optimize bottlenecks\n     - Implement caching strategies\n     - Optimize data loading and preprocessing\n     - Memory management for large-scale training\n\n  3. Testing framework:\n     - Unit tests for data transformations\n     - Integration tests for pipeline components\n     - Model quality tests (invariance, directional)\n     - Performance regression tests\n\n  Deliver production-ready, maintainable code with full test coverage.\n\u003C\u002FTask>\n\n## Phase 3: Production Deployment & Serving\n\n\u003CTask>\nsubagent_type: mlops-engineer\nprompt: |\n  Design production deployment for models from: {phase2.ml-engineer.output}\n  With optimized code from: {phase2.python-pro.output}\n\n  Implementation requirements:\n  1. Model serving infrastructure:\n     - REST\u002FgRPC APIs with FastAPI\u002FTorchServe\n     - Batch prediction pipelines (Airflow\u002FKubeflow)\n     - Stream processing (Kafka\u002FKinesis integration)\n     - Model serving platforms (KServe\u002FSeldon Core)\n\n  2. Deployment strategies:\n     - Blue-green deployments for zero downtime\n     - Canary releases with traffic splitting\n     - Shadow deployments for validation\n     - A\u002FB testing infrastructure\n\n  3. CI\u002FCD pipeline:\n     - GitHub Actions\u002FGitLab CI workflows\n     - Automated testing gates\n     - Model validation before deployment\n     - ArgoCD for GitOps deployment\n\n  4. Infrastructure as Code:\n     - Terraform modules for cloud resources\n     - Helm charts for Kubernetes deployments\n     - Docker multi-stage builds for optimization\n     - Secret management with Vault\u002FSecrets Manager\n\n  Provide complete deployment configuration and automation scripts.\n\u003C\u002FTask>\n\n\u003CTask>\nsubagent_type: kubernetes-architect\nprompt: |\n  Design Kubernetes infrastructure for ML workloads from: {phase3.mlops-engineer.output}\n\n  Kubernetes-specific requirements:\n  1. Workload orchestration:\n     - Training job scheduling with Kubeflow\n     - GPU resource allocation and sharing\n     - Spot\u002Fpreemptible instance integration\n     - Priority classes and resource quotas\n\n  2. Serving infrastructure:\n     - HPA\u002FVPA for autoscaling\n     - KEDA for event-driven scaling\n     - Istio service mesh for traffic management\n     - Model caching and warm-up strategies\n\n  3. Storage and data access:\n     - PVC strategies for training data\n     - Model artifact storage with CSI drivers\n     - Distributed storage for feature stores\n     - Cache layers for inference optimization\n\n  Provide Kubernetes manifests and Helm charts for entire ML platform.\n\u003C\u002FTask>\n\n## Phase 4: Monitoring & Continuous Improvement\n\n\u003CTask>\nsubagent_type: observability-engineer\nprompt: |\n  Implement comprehensive monitoring for ML system deployed in: {phase3.mlops-engineer.output}\n  Using Kubernetes infrastructure: {phase3.kubernetes-architect.output}\n\n  Monitoring framework:\n  1. Model performance monitoring:\n     - Prediction accuracy tracking\n     - Latency and throughput metrics\n     - Feature importance shifts\n     - Business KPI correlation\n\n  2. Data and model drift detection:\n     - Statistical drift detection (KS test, PSI)\n     - Concept drift monitoring\n     - Feature distribution tracking\n     - Automated drift alerts and reports\n\n  3. System observability:\n     - Prometheus metrics for all components\n     - Grafana dashboards for visualization\n     - Distributed tracing with Jaeger\u002FZipkin\n     - Log aggregation with ELK\u002FLoki\n\n  4. Alerting and automation:\n     - PagerDuty\u002FOpsgenie integration\n     - Automated retraining triggers\n     - Performance degradation workflows\n     - Incident response runbooks\n\n  5. Cost tracking:\n     - Resource utilization metrics\n     - Cost allocation by model\u002Fexperiment\n     - Optimization recommendations\n     - Budget alerts and controls\n\n  Deliver monitoring configuration, dashboards, and alert rules.\n\u003C\u002FTask>\n\n## Configuration Options\n\n- **experiment_tracking**: mlflow | wandb | neptune | clearml\n- **feature_store**: feast | tecton | databricks | custom\n- **serving_platform**: kserve | seldon | torchserve | triton\n- **orchestration**: kubeflow | airflow | prefect | dagster\n- **cloud_provider**: aws | azure | gcp | multi-cloud\n- **deployment_mode**: realtime | batch | streaming | hybrid\n- **monitoring_stack**: prometheus | datadog | newrelic | custom\n\n## Success Criteria\n\n1. **Data Pipeline Success**:\n   - \u003C 0.1% data quality issues in production\n   - Automated data validation passing 99.9% of time\n   - Complete data lineage tracking\n   - Sub-second feature serving latency\n\n2. **Model Performance**:\n   - Meeting or exceeding baseline metrics\n   - \u003C 5% performance degradation before retraining\n   - Successful A\u002FB tests with statistical significance\n   - No undetected model drift > 24 hours\n\n3. **Operational Excellence**:\n   - 99.9% uptime for model serving\n   - \u003C 200ms p99 inference latency\n   - Automated rollback within 5 minutes\n   - Complete observability with \u003C 1 minute alert time\n\n4. **Development Velocity**:\n   - \u003C 1 hour from commit to production\n   - Parallel experiment execution\n   - Reproducible training runs\n   - Self-service model deployment\n\n5. **Cost Efficiency**:\n   - \u003C 20% infrastructure waste\n   - Optimized resource allocation\n   - Automatic scaling based on load\n   - Spot instance utilization > 60%\n\n## Final Deliverables\n\nUpon completion, the orchestrated pipeline will provide:\n- End-to-end ML pipeline with full automation\n- Comprehensive documentation and runbooks\n- Production-ready infrastructure as code\n- Complete monitoring and alerting system\n- CI\u002FCD pipelines for continuous improvement\n- Cost optimization and scaling strategies\n- Disaster recovery and rollback procedures\n",{"name":5,"risk":32,"source":29,"date_added":38,"description":7},"2026-02-27","main","d8518247cb8ab8b2dac57a6be33543ffd23b97c5","data-ml\u002Fsickn33\u002Fmachine-learning-ops-ml-pipeline",0,"https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry","https:\u002F\u002Fgithub.com\u002Fmattonweb","2026-05-29 08:42:23",{"source":8,"skillName":5,"commands":47,"prompt":57,"zipAvailable":58},[48,51,54],{"tool":49,"command":50},"npx","npx skills add mattonweb\u002Fai-assistant-foundry --skill machine-learning-ops-ml-pipeline",{"tool":52,"command":53},"bunx","bunx skills add mattonweb\u002Fai-assistant-foundry --skill machine-learning-ops-ml-pipeline",{"tool":55,"command":56},"pnpm","pnpm dlx skills add mattonweb\u002Fai-assistant-foundry --skill machine-learning-ops-ml-pipeline","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fdata-ml\u002Fsickn33\u002Fmachine-learning-ops-ml-pipeline\u002FSKILL.md",true,[],[61,97,111,129,150,176],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":77,"publishedAt":33},"01M6P02DENRHN6KVGYC55FKASJ","writing-clearly-and-concisely","Use when writing prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Applies Strunk's timeless rules for clearer, stronger, more professional writing.","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fwrite-technical\u002Fsoftaworks\u002Fwriting-clearly-and-concisely\u002FSKILL.md",[67,72],{"categoryNo":68,"categoryName":69,"taxonomyType":70,"taxonomyTypeDesc":71},"01JBSEEDCAT000000000000018","写作与文案","SCENARIO","使用场景",{"categoryNo":73,"categoryName":74,"taxonomyType":75,"taxonomyTypeDesc":76},"01JBSEEDCAT000000000000036","内容创作者","PROFESSION","适用职业",[78,79,80,81,31,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96],"writing clearly and concisely","writing","concisely","text","more","timeless","error","professional","prose","reports","use","read","Rules","commit messages","messages","error messages","ui","documentation","commit",{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":33},"01M6P0286HZ8RWMFCCPC8HF7SC","crafting-effective-readmes","Use when writing or improving README files. Not all READMEs are the same — provides templates and guidance matched to your audience and project type.","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fwrite-technical\u002Fsoftaworks\u002Fcrafting-effective-readmes\u002FSKILL.md",[],[104,105,79,106,107,88,108,109,110],"crafting effective readmes","README","templates","type","project","guidance","files",{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":120,"publishedAt":33},"01M6P025A41BBTPWVTZ3GFQVG6","wiki-architect","You are a documentation architect that produces structured wiki catalogues and onboarding guides from codebases.","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fwrite-technical\u002Fsickn33\u002Fwiki-architect\u002FSKILL.md",[117],{"categoryNo":118,"categoryName":119,"taxonomyType":75,"taxonomyTypeDesc":76},"01JBSEEDCAT000000000000038","人力资源",[121,122,123,32,124,125,126,127,29,128,95],"wiki","architect","wiki architect","you","documentation architect","guides","structured","onboarding",{"skillNo":130,"name":131,"nameZh":6,"description":132,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":133,"categories":134,"keywords":135,"publishedAt":33},"01M6P01NSKPXP77NWB4K0PXRK5","readme","You are an expert technical writer creating comprehensive project documentation. Your goal is to write a README.md that is absurdly thorough—the kind of documentation you wish every project had.","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fwrite-technical\u002Fsickn33\u002Freadme\u002FSKILL.md",[],[105,136,137,138,124,139,140,108,141,142,39,143,144,145,146,147,95,148,149],"project documentation","tree","kind","expert","github","wish","writer","goal","write","technical writer","technical","https","write a readme","safe",{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":156,"publishedAt":33},"01M6P01GP0NEN8KT3A3SA6ZFVE","ubiquitous-language","Extract a DDD-style ubiquitous language glossary from the current conversation, flagging ambiguities and proposing canonical terms. Saves to UBIQUITOUS_LANGUAGE.md. Use when user wants to define domain terms, build a glossary, harden terminology, create a ubiquitous language, or mentions \"domain model\" or \"DDD\".","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fwrite-technical\u002Fmattpocock\u002Fubiquitous-language\u002FSKILL.md",[],[157,158,88,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175],"ubiquitous language","language","extract","user","glossary","model","terms","ddd","create","build","true","canonical","style","define","harden","domain model","domain","terminology","conversation",{"skillNo":177,"name":178,"nameZh":6,"description":179,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":180,"categories":181,"keywords":184,"publishedAt":33},"01M6P01BJBP9XP4TDRT7QD1ZVA","technical-writing-style","Enforce clear, direct writing in software development documents including OKRs, PRDs, tech specs, architecture decision records (ADRs), git commit messages, PR descriptions, READMEs, changelogs, runbooks, incident postmortems, API docs, and onboarding guides. Trigger whenever a user asks to write, edit, revise, or review any of these document types. Also trigger when the user says writing \"sounds like AI\", \"feels robotic\", \"needs polish\", or asks to \"make it sound more natural\" in the context of technical or professional prose. Do not trigger for fiction, poetry, creative writing, academic papers, or legal documents unless the user explicitly asks for a more natural tone.","https:\u002F\u002Fgithub.com\u002Fmattonweb\u002Fai-assistant-foundry\u002Fblob\u002Fmain\u002Fwrite-technical\u002Fmattnebr\u002Ftechnical-writing-style\u002FSKILL.md",[182,183],{"categoryNo":68,"categoryName":69,"taxonomyType":70,"taxonomyTypeDesc":71},{"categoryNo":73,"categoryName":74,"taxonomyType":75,"taxonomyTypeDesc":76},[185,186,169,146,79,187,85,188,189,190,191,192,193,194,195,92,196,197,96,198],"writing style","technical writing style","technical writing","papers","types","runbooks","software development","incident","Poetry","decision","api","documents","architecture","review"]