aigroup-workflow

Build Recovery Team 修复构建/类型/依赖/配置错误，不建 session

主会话现在需要创建一个 Git 提交。

初始化项目 AI 上下文 — 根 CLAUDE.md 只做导航，详细内容进 docs/

启动轻量 Planning Team 输出实施计划，不建 session

启动轻量 Review Team 审查当前改动，不建 session

You are a Senior Accessibility Architect. Your goal is to ensure that every digital product is Perceivable, Operable, Understandable, and Robust (POUR) for all users, including those with visual, auditory, motor, or cognitive disabilities.

Software architecture specialist for system design, scalability, and technical decision-making. Use PROACTIVELY when planning new features, refactoring large systems, or making architectural decisions.

Build and TypeScript error resolution specialist. Use PROACTIVELY when build fails or type errors occur. Fixes build/type errors only with minimal diffs, no architectural edits. Focuses on getting the build green quickly.

Personal communication chief of staff that triages email, Slack, LINE, and Messenger. Classifies messages into 4 tiers (skip/info_only/meeting_info/action_required), generates draft replies, and enforces post-send follow-through via hooks. Use when managing multi-channel communication workflows.

Designs feature architectures by analyzing existing codebase patterns and conventions, then providing implementation blueprints with concrete files, interfaces, data flow, and build order.

Use when creating new skills, editing existing skills, or verifying skills work before deployment

Use when fine-tuning LLMs, training custom models, or adapting foundation models for specific tasks. Invoke for configuring LoRA/QLoRA adapters, preparing JSONL training datasets, setting hyperparameters for fine-tuning runs, adapter training, transfer learning, finetuning with Hugging Face PEFT, OpenAI fine-tuning, instruction tuning, RLHF, DPO, or quantizing and deploying fine-tuned models. Trigger terms include: LoRA, QLoRA, PEFT, finetuning, fine-tuning, adapter tuning, LLM training, model training, custom model.

Designs and implements production-grade ML pipeline infrastructure: configures experiment tracking with MLflow or Weights & Biases, creates Kubeflow or Airflow DAGs for training orchestration, builds feature store schemas with Feast, deploys model registries, and automates retraining and validation workflows. Use when building ML pipelines, orchestrating training workflows, automating model lifecycle, implementing feature stores, managing experiment tracking systems, setting up DVC for data versioning, tuning hyperparameters, or configuring MLOps tooling like Kubeflow, Airflow, MLflow, or Prefect.

Designs and implements production-grade RAG systems by chunking documents, generating embeddings, configuring vector stores, building hybrid search pipelines, applying reranking, and evaluating retrieval quality. Use when building RAG systems, vector databases, or knowledge-grounded AI applications requiring semantic search, document retrieval, context augmentation, similarity search, or embedding-based indexing.

Use when writing Spark jobs, debugging performance issues, or configuring cluster settings for Apache Spark applications, distributed data processing pipelines, or big data workloads. Invoke to write DataFrame transformations, optimize Spark SQL queries, implement RDD pipelines, tune shuffle operations, configure executor memory, process .parquet files, handle data partitioning, or build structured streaming analytics.