kimtth

Plan, implement, review, and troubleshoot Microsoft Fabric ALM and CI/CD workflows using Git integration, deployment pipelines, Variable Libraries, Fabric REST APIs, fabric-cicd, GitHub Actions, or Azure DevOps. Use when the user asks about source control, deploy, promote, release, dev/test/prod, environment variables, deployment pipeline automation, Git sync, fabric-cicd, or Fabric item definition validation.

Design, configure, evaluate, and govern Microsoft Fabric Data Agents for natural-language Q&A over Lakehouse, Warehouse, Power BI semantic model, KQL database, mirrored database, ontology, or Microsoft Graph data. Use when the user asks for Fabric data agent, conversational analytics, NL2SQL, NL2DAX, NL2KQL, data-agent instructions, example queries, agent evaluation, publishing, sharing, governance, diagnostics, or ALM.

Fabric Lakehouse design, schemas, shortcuts, security, optimization, and PySpark patterns. Use when designing Lakehouse solutions, managing Delta tables, configuring OneLake shortcuts, or writing PySpark/Spark SQL code for Fabric notebooks.

Develop Microsoft Fabric Spark/data engineering workflows and write code in Fabric Notebook cells with intelligent routing to specialized resources. Provides workspace/lakehouse management, notebook code authoring (PySpark, Scala, SparkR, SQL), and routes to: data engineering patterns, development workflow, or infrastructure orchestration. Use when the user wants to: (1) manage Fabric workspaces and resources, (2) write or debug code in notebook cells, (3) use notebookutils, (4) develop notebooks and PySpark applications, (5) design data pipelines, (6) provision infrastructure as code. Triggers: "develop notebook", "data engineering", "workspace setup", "pipeline design", "infrastructure provisioning", "Delta Lake patterns", "Spark development", "lakehouse configuration", "write notebook code", "notebookutils", "notebook cell", "PySpark notebook", "%%sql cell", "%%configure", "fabric notebook", "run notebook", "notebook deployment".

Diagnose failed Spark jobs, unhealthy Livy sessions, and performance bottlenecks in Microsoft Fabric via read-only CLI triage. Use when the user wants to: (1) diagnose why a Spark job, notebook run, or Lakehouse job failed, (2) triage stuck or dead Livy sessions, (3) identify OOM, shuffle spill, or data skew, (4) retrieve driver and executor logs or Spark Advisor findings, (5) copy event logs and start a local Spark History Server, (6) diagnose all Spark activities within a failed pipeline run. Triggers: "diagnose my failed notebook", "why did my spark job fail", "triage spark failure", "diagnose pipeline run failure", "why did my pipeline fail", "livy session stuck in starting", "spark executor OOM", "check spark advisor findings", "shuffle spill diagnosis", "why did my lakehouse job fail", "diagnose lakehouse table load", "data skew diagnosis", "open spark history server locally", "analyze spark failure logs", "spark job triage".

Discover Fabric APIs, OpenAPI specs, item schemas, and best practices using the Fabric MCP Server. Use when exploring available Fabric workloads, looking up API specifications, finding item definition formats, or managing OneLake files programmatically. All MCP tools run locally for reference.

Core Microsoft Fabric platform reference: topology, authentication, token scopes, REST API base URL, pagination, long-running operations, throttling, workspace and item resolution, OneLake access, and common gotchas. Use this skill whenever working with Fabric REST APIs, managing workspaces/items, or troubleshooting auth errors.

Look up Microsoft API references, find working code samples, and verify SDK code is correct. Use when working with Azure SDKs, .NET libraries, or Microsoft APIs to find the right method, check parameters, get working examples, or troubleshoot errors. Catches hallucinated methods, wrong signatures, and deprecated patterns by querying official docs.

Workflow and tools for adding new entries from temp.md to the section files. Includes legend format, section reference, code tools, and common pitfalls. USE FOR: Adding new resources to the knowledge base. DO NOT USE FOR: Editing existing entries or restructuring sections.

Workflow for updating the LLM landscape paper pool (section/x_llm_papers.md) using fetch_llm_papers.py. Covers full re-fetch, resume from checkpoint, and adding new topics. USE FOR: Refreshing citation counts, expanding topic coverage. DO NOT USE FOR: Adding hand-curated entries to section files (use add-new-entry), updating RAG/Agent citation sections in best_practices.md (use update-cite-count).

Workflow for updating the popular LLM applications pool (section/x_llm_apps.md) using get_app_list_by_github_star.py. Covers full refresh, alternate exports, topic tuning, and common pitfalls. USE FOR: Refreshing the ranked GitHub applications list linked from applications.md. DO NOT USE FOR: Hand-curating application entries inside applications.md or adding GitHub star badges to the generated file.

Generate synthetic and simulated datasets for evaluation and fine-tuning using Azure AI Foundry simulators. Create non-adversarial task data, adversarial safety data, and conversation datasets without manual data collection.

Train or fine-tune LLMs on Azure ML managed compute with TRL trainers. Uses direct trainer loops (SFT, DPO, RL) without relying on serverless APIs or Hugging Face infrastructure.

Evaluate generative AI applications and models locally or in the cloud using Azure AI Evaluation SDK. Measure quality, safety, and performance with built-in and custom evaluators.