Import and use Kubb's generated code (types, clients, hooks, schemas, mocks). Use when writing app or test code that consumes a Kubb build.

Reviews pull requests against project standards and posts review comments via the gh CLI. Use when reviewing PRs, checking coding standards compliance, or performing approval reviews.

Facade-specific YARD documentation rules for Git::Repository::* topic modules and their facade methods, overriding and extending the general yard-documentation skill. Use when writing or reviewing YARD docs for facade modules under lib/git/repository/.

Conventions for writing and reviewing unit and integration tests for Git::Commands::* classes. Use when scaffolding new command tests or auditing existing ones.

Pose classification using ST-GCN (Spatial Temporal Graph Convolutional Network). Classifies skeleton sequences into action categories from pose-keypoint data. Use when training, evaluating, exporting, or running inference for a TAO pose-classification model. Trigger phrases include "train pose classification", "skeleton action recognition", "ST-GCN", "keypoint sequence classifier".

NVDINOv2 for self-supervised visual representation learning. Trains vision transformers via self-distillation (teacher-student) without labels and produces general-purpose visual features. Use when training, exporting, or running inference for a TAO NVDINOv2 backbone. Trigger phrases include "train NVDINOv2", "self-supervised ViT pretraining", "DINOv2 backbone", "visual representation learning".

Masked Auto-Encoder (MAE) for self-supervised pretraining and fine-tuning. Masks random patches and reconstructs them to learn visual representations; supports pretrain and finetune stages. Use when training, evaluating, exporting, or running inference for a TAO MAE backbone. Trigger phrases include "pretrain MAE", "self-supervised vision pretraining", "Masked Autoencoder", "Mask Auto-Encoder", "MAE fine-tune".

Action recognition from video sequences. Supports RGB, optical flow, and joint (multi-stream) input types for classifying temporal actions in video clips. Use when training, evaluating, exporting, or running inference on a TAO action-recognition model. Trigger phrases include "train action recognition", "video action classification", "RGB + optical flow action model", "TAO ActionRecognition".

Extract false-positive and false-negative gaps from VLM binary-classification-question (BCQ, yes/no) predictions. Use when the user asks to "analyze VLM BCQ gaps", "extract VLM false positives and false negatives", or identify failure cases from a predictions JSON for DEFT root-cause analysis on a binary-classification VLM workflow.

Editing conventions for files under `.claude/skills/**` (markdown style, frontmatter format, SKILL.md vs `references/` split, upstream-vendored skills). Use when creating a new skill, editing any `SKILL.md` or `references/*.md` file, refactoring a skill description, or after adding/removing a skill.

How to work in a Plain Notes project (the `plain-notes` starter pack): a flat notes/ folder plus a daily/ journal. The 'I just want to write' layout. Read when the project has these folders, OR when asked to jot a note, capture a quick thought, or write today's journal entry. Carries the linking habit and daily-entry behavior so templates and folder descriptions stay minimal. Complements the platform `open-knowledge` skill; does not replace it.

Investigate a topic against preserved sources and write a draft-status research article under `research/` in a Knowledge Base project (the `knowledge-base` starter pack). Read when asked to research a topic, compare options, synthesize sources, gather evidence, or extend an existing research doc. Carries the full procedure: scan existing coverage, agree a research rubric, capture every source verbatim before analyzing, write the article incrementally so a crash never loses work, cite every claim, and link it back into the graph. Does not promote findings to canonical knowledge — that is the sibling `consolidate-notes` skill, after a decision lands.

Write or refine Chinese technical notes and articles in a concise, evidence-led voice. Use for substantive drafting or rewriting, not every Chinese reply.

Guidance on selecting the appropriate test suite (Unit, API, or E2E tests) when writing new tests in Chrome DevTools. MUST be used when writing new tests.

Review and validate claims using counter-hypothesis testing. Use when verifying technical content, checking references, validating API endpoints, or performing quality assurance on deliverables.

Review and validate claims using counter-hypothesis testing. Use when verifying technical content, checking references, validating API endpoints, or performing quality assurance on deliverables.

**WORKFLOW SKILL** — Prepares Azure.ResourceManager SDK refresh pull requests in azure-sdk-for-net. WHEN: "prepare sdk refresh", "refresh Azure.ResourceManager package", "update ARM SDK from autorest tag", "refresh changelog dependencies". INVOKES: git and GitHub pull request tools for branch, commit, push, and PR creation. FOR SINGLE OPERATIONS: Use git or GitHub tools directly.

Read before adding or changing any front/ UI element - a control, layout, button, or interaction pattern. Covers the don't-invent-new-UX-without-a-PRD rule and the priority order for design tradeoffs (existing behavior > intuitiveness > information density > usability > utility > uniqueness > industry practices > generic UI).

Rigorous PRD-writing methodology for NetAlertX — challenge the idea, verify every claim against actual code, trace every downstream consumer of a mechanism, evaluate performance impact against the schema/indexes that actually exist, record rejected alternatives and open-issue decisions explicitly, and do a dedicated final-check pass before calling it done. Use this when asked to write, draft, or review a PRD, design doc, or feature proposal.

Explore and understand Nx workspaces. USE WHEN answering questions about the workspace, projects, or tasks. ALSO USE WHEN an nx command fails or you need to check available targets/configuration before running a task. EXAMPLES: 'What projects are in this workspace?', 'How is project X configured?', 'What depends on library Y?', 'What targets can I run?', 'Cannot find configuration for task', 'debug nx task failure'.