Debugging

Handles LLM-as-judge evaluation workflows on Arize including creating/updating evaluators, running evaluations on spans or experiments, managing tasks, trigger-run operations, column mapping, and continuous monitoring. Use when the user mentions create evaluator, LLM judge, hallucination, faithfulness, correctness, relevance, run eval, score spans, score experiment, trigger-run, column mapping, continuous monitoring, or improve evaluator prompt.

Convert .NET projects and solutions (.sln, .slnx) to NuGet Central Package Management (CPM) using Directory.Packages.props. USE FOR: converting to CPM, centralizing or aligning NuGet package versions across multiple projects, inlining MSBuild version properties from Directory.Build.props into Directory.Packages.props, resolving version conflicts or mismatches across a solution or repository, updating or bumping or syncing package versions across projects. Also activate when packages are out of sync, drifting, or inconsistent -- even without the user mentioning CPM. Provides baseline build capture, version conflict resolution, build validation with binlog comparison, and a structured post-conversion report. DO NOT USE FOR: packages.config projects (must migrate to PackageReference first) or repositories that already have CPM fully enabled.

Standard single-cell RNA-seq analysis pipeline. Use for QC, normalization, dimensionality reduction (PCA/UMAP/t-SNE), clustering, differential expression, visualization, and converting R-friendly single-cell formats such as Seurat or SingleCellExperiment RDS files into h5ad for Scanpy. Best for exploratory scRNA-seq analysis with established workflows. For deep learning models use scvi-tools; for data format questions use anndata.

Semi-automated design quality review for Flows apps. Runs concrete repo probes (grep, lint, build) to propose a draft 1–5 score for each of the official 10 quality-guidelines questions from docs.cognite.com/cdf/flows/guides/quality-guidelines, then asks the user to confirm or override each score. Still requires the user to walk their tasks end-to-end in the running app (Step 2) since navigation and clickability feel cannot be measured statically. Writes reviews/design-review/feedback-round-<N>/design-review-report.md with an overall average and prioritized fix lists. Use when the user asks to run a Flows design review, run the design quality assessment, or run flows-design-review. Must be run AFTER flows-code-review reaches 0 Must Fix and BEFORE flows-external-app-submit.

Detects MSBuild projects with conflicting OutputPath or IntermediateOutputPath. USE FOR: builds failing with 'Cannot create a file when that file already exists', 'The process cannot access the file because it is being used by another process', intermittent build failures that succeed on retry, or missing/overwritten outputs in multi-project or multi-targeting builds where bin/obj (or project.assets.json) collide. Common causes: shared OutputPath, missing AppendTargetFrameworkToOutputPath, extra global properties (e.g. PublishReadyToRun), or SetTargetFramework on a ProjectReference to a single-targeting project. DO NOT USE FOR: file access errors unrelated to MSBuild (OS-level locking), single-project single-TFM builds, non-MSBuild build systems.

Build and troubleshoot TorchDrug 0.2.1 workflows for molecular graphs, property prediction, self-supervised pretraining, molecule generation, retrosynthesis, protein representation learning, and knowledge graph reasoning. Use when code imports torchdrug or needs its datasets, models, tasks, or Engine.