Play .cast recordings in iTerm2 with speed controls. TRIGGERS - play recording, asciinema play, view cast.

Start terminal recording with asciinema. TRIGGERS - record session, capture terminal, start recording.

Production wiring for the MiniMax-M3 model — empirically verified flags, capabilities, and limits (thinking control via reasoning_split, native vision, response_format, 512K input ceiling, 512K output cap, n=1, docs-vs-reality discrepancies). Use when wiring or tuning MiniMax-M3, choosing M3 vs M2.7/-highspeed, switching a service off M2.7-highspeed onto M3, getting clean output without <think>, or asking what M3 supports / how big its context is. TRIGGERS - MiniMax M3, MiniMax-M3, M3 model, switch to M3, reasoning_split, M3 context length, M3 vision, M3 options, get the most out of M3.

Operate and troubleshoot the MT5 tick collection system on Linux/Wine. Systemd topology, gap detection, restart recovery, daily.

Machine-wide loop muster — enumerate every registered loop with health, ownership, and staleness. Renamed from 'status' to avoid.

Mark LOOP_CONTRACT.md completed, append DONE section, send PushNotification, let loop terminate naturally.

user wants to archive, dump, or back up an entire Telegram channel or chat history to NDJSON with all media files downloaded. Full history.

Interactive chezmoi drift check and sync. TRIGGERS - chezmoi sync, sync dotfiles, dotfile drift, chezmoi guard, chezmoi check.

user wants to clean up Telegram by purging deleted or ghost accounts from their dialog list and contacts, or removing spam ghosts that clutter.

user wants to list all Telegram chats, groups, and channels, see their contacts, find a chat ID, browse conversations, or check account info via.

user wants to manage Telegram group or channel members, including inviting users, kicking or banning someone, listing members, or filtering.

user wants to search for messages across all Telegram chats or within a specific chat, find old messages by text, or look up Telegram message.

Triage autoloop fleet health. Reports GREEN/YELLOW/RED per loop with remediation hints. Renamed from 'doctor' to avoid clashing.

Use when user wants to create a new Telegram group, supergroup, or channel, optionally inviting members on creation.

Use when user wants to delete, remove, or unsend Telegram messages from a chat, either for everyone or just for themselves.

Use when user wants to download, save, or extract media files such as photos, videos, and documents from Telegram messages.

Use when user wants to find, look up, or resolve a Telegram user by username, phone number, or ID, or get someone's Telegram profile info.

Build FloatingClock from source and install to /Applications, then launch. Use when user wants to install the floating-clock plugin's macOS app.

user wants to mark Telegram chats as read, clear unread badges and mentions, dismiss notifications, or acknowledge messages to remove the unread.

Use when user wants to pin or unpin a message in a Telegram chat, group, or channel, or manage pinned messages.

Show openwolf health for cwd (files, hooks, ledger, anatomy, daemon). TRIGGERS - openwolf status, openwolf health.

Quit FloatingClock, remove it from /Applications, and clear its saved preferences. Use when the user wants to completely uninstall the.

Send a Pushover EMERGENCY (priority 2) alert that repeats until acknowledged, then poll the receipt until the user acks or it expires. Use when something is urgent and must not be missed - production breakage, a blocked autonomous loop needing immediate input, an alarm-until-acknowledged notification. For routine notifications use send-notification. TRIGGERS - emergency alert, urgent pushover, alarm until acknowledged, priority 2, wake me up.

Launch FloatingClock. Prefers the installed /Applications version; falls back to the local build. Use when the user wants to open the.

Complete FFmpeg command syntax reference covering option ordering, input vs output options, stream specifiers, and position-sensitive options. PROACTIVELY activate for: (1) Command syntax questions, (2) Option placement issues, (3) Input vs output option confusion, (4) Stream specifier syntax, (5) -ss/-t/-to position questions, (6) Global vs per-file options, (7) Multiple input/output handling, (8) Option order errors. Provides: Correct option placement rules, input-only vs output-only options, position-sensitive option behavior, stream specifier syntax, common mistakes and fixes.

ADF activity nesting rules, limits, and validation gotchas. PROACTIVELY activate for: (1) ADF activity nesting rules (ForEach inside If, Switch, Until), (2) ForEach limitations (no nested ForEach, item limits), (3) pipeline validation errors and ARM-template build failures, (4) linked service authentication issues, (5) Set Variable restrictions in ForEach, (6) Data Flow constraints (transformation limits, source/sink rules), (7) parameter and variable scope, (8) max activities per pipeline, (9) pipeline concurrency limits. Provides: nesting rule matrix, activity limit reference, validation error catalog, and refactoring patterns for unsupported nesting.

Complete karaoke subtitle system and advanced animated text effects. PROACTIVELY activate for: (1) Karaoke-style highlighted lyrics, (2) ASS/SSA advanced subtitle styling, (3) Scrolling credits (horizontal/vertical), (4) Typewriter text animation, (5) Bouncing/moving text, (6) Text fade in/out effects, (7) Word-by-word text reveal, (8) Kinetic typography, (9) Lower thirds animation, (10) Countdown timers and dynamic text. Provides: ASS karaoke timing format, drawtext with time expressions, scrolling text patterns, text animation formulas, kinetic typography techniques, subtitle styling reference, multi-line animated text.

Complete Python asyncio system. PROACTIVELY activate for: (1) async/await syntax, (2) asyncio.gather for concurrent execution, (3) TaskGroup (3.11+), (4) Semaphores for rate limiting, (5) Timeouts with asyncio.timeout, (6) Producer-consumer with Queue, (7) Async generators and context managers, (8) uvloop performance, (9) Common async gotchas. Provides: Concurrent patterns, I/O optimization, async libraries (aiohttp, httpx, asyncpg). Ensures correct async patterns without blocking.

Complete Python type hints system. PROACTIVELY activate for: (1) Built-in generics (list[str], dict[str, int]), (2) Union types (str | None), (3) Type parameter syntax 3.12+, (4) Protocol for structural typing, (5) TypedDict for dict schemas, (6) Literal and Final types, (7) TypeGuard and TypeIs, (8) ParamSpec for decorators, (9) Mypy/Pyright configuration. Provides: Type syntax, Protocol patterns, TypedDict, mypy config. Ensures static type safety with gradual typing strategy.

ADF + Microsoft Fabric / OneLake 2025 integration. PROACTIVELY activate for: (1) Fabric Lakehouse connector in ADF, (2) Fabric Warehouse connector in ADF, (3) OneLake shortcuts and cross-workspace data, (4) Invoke Pipeline activity for cross-platform orchestration (ADF -> Fabric, Fabric -> ADF), (5) copying data between ADF and Microsoft Fabric workspaces, (6) authenticating with Fabric workspace identity, (7) cross-platform parameter passing, (8) hybrid ADF + Fabric pipeline patterns. Provides: Fabric connector setup, Invoke Pipeline templates, OneLake shortcut patterns, and ADF-to-Fabric migration guidance.

Parallel and concurrent processing patterns in bash including GNU Parallel, xargs -P, job pools, and async patterns (2025). PROACTIVELY activate for: (1) running tasks in parallel from bash, (2) xargs -P for parallel command execution, (3) GNU parallel for complex distribution, (4) background jobs with & and wait, (5) implementing a concurrency cap (job pool of N workers), (6) collecting exit codes from parallel children, (7) preventing race conditions in shared output files, (8) parallel file processing across many cores. Provides: xargs -P recipes, GNU parallel quick reference, job-pool implementations in pure bash, exit-code aggregation patterns, and FIFO-based work queues.

ADF + Databricks 2025 integration patterns. PROACTIVELY activate for: (1) Databricks Job activity in ADF, (2) DatabricksJob (preview) vs DatabricksNotebook activity, (3) ServiceNow V2 connector, (4) ADF managed identity authentication for Databricks, (5) Databricks serverless linked services, (6) Snowflake V2 connector, (7) Databricks job parameters and outputs, (8) MFA enforcement and authentication updates, (9) Unity Catalog integration, (10) Delta Live Tables orchestration from ADF. Provides: Databricks linked service templates (PAT, MSI, serverless), DatabricksJob activity examples, parameter passing recipes, and authentication migration guidance.

Process substitution, named pipes (FIFOs), and advanced IPC patterns for efficient bash data streaming (2025). PROACTIVELY activate for: (1) using process substitution with <(cmd) and >(cmd), (2) creating named pipes (mkfifo) for inter-process streaming, (3) avoiding intermediate temp files in pipelines, (4) tee-ing into multiple consumers, (5) feeding diff/comm with command output, (6) building shell-only producers and consumers, (7) reading from a FIFO with timeout, (8) signal-safe FIFO cleanup. Provides: process substitution patterns, mkfifo recipes, multi-consumer tee patterns, and signal-safe cleanup with trap.

This skill should be used to create or update a C4 architecture diagram alongside an ADR, especially in LikeC4. PROACTIVELY activate on "add a C4 diagram", "create a context diagram", "container view", "LikeC4 model", "draw the architecture", "diagram this decision", "architecture diagram for the ADR", "visualize the components", "C4 view", or "Simon Brown C4." Provides: canonical-C4 Context and Container views from confirmed elements, relationship descriptions, and strict refusal of non-canonical diagram types.

This skill should be used before drafting an ADR when context, components, relationships, deciders, or prior decisions are unclear. PROACTIVELY activate on "I want to write an ADR for X", "gather context for a decision", "discovery for an ADR", "ADR intake", "ADR pre-flight", "prep work before an ADR", "what's the current architecture?", or "set up a decision log." Provides: zero-hallucination Q&A, human-confirmed discovery brief, open-questions register, and five required confirmations before drafting.

This skill should be used when the user asks about terminal emulator compatibility, operating-system differences, tmux/screen/Zellij, SSH/remote terminals, Windows ConPTY/classic console/Windows Terminal, Unix PTYs, macOS Terminal/iTerm2, Kitty/Alacritty/WezTerm/foot/GNOME Terminal/Konsole, terminal feature matrices, escape-sequence passthrough, latency, resize propagation, or graceful degradation. PROACTIVELY activate for: compatibility audits, emulator bugs, remote TUI performance, multiplexer passthrough. Provides: support matrices, platform caveats, and fallback rules.

Complete color manipulation and green screen effects system. PROACTIVELY activate for: (1) Green screen/chromakey removal, (2) Color grading with LUTs, (3) Color correction (levels, curves, white balance), (4) Colorkey/color removal effects, (5) Hue/saturation manipulation, (6) Color space conversions (BT.709, BT.2020, HDR), (7) Color isolation effects, (8) Teal and orange look, (9) Vintage/film looks, (10) Color keying for transparency. Provides: chromakey and colorkey filters, LUT application (lut3d), curves and levels adjustment, color balance, selective color manipulation, color space handling, HDR tone mapping, professional color grading chains.

Advanced bash array patterns including indexed and associative arrays, mapfile/readarray, slicing, and array manipulation (2025). PROACTIVELY activate for: (1) using bash arrays effectively, (2) reading file contents into an array via mapfile/readarray, (3) authoring associative arrays (declare -A), (4) iterating arrays safely under set -u, (5) deleting/inserting array elements, (6) joining and splitting arrays, (7) passing arrays to functions via nameref (declare -n), (8) array slicing and parameter expansion, (9) sorting and deduplicating arrays in pure bash. Provides: indexed and associative array reference, mapfile and readarray recipes, nameref-based pass-by-reference, slicing/expansion patterns, and gotchas for unset variables and word splitting.

Bash 5.3 new features and modern patterns (2025). PROACTIVELY activate for: (1) Bash 5.3 specific features (BASH_TRAPSIG, in-shell command substitution, ${|} REPLY syntax), (2) checking bash version compatibility, (3) migrating scripts to take advantage of 5.3 additions, (4) C23 conformance changes in bash 5.3, (5) new shopt and bind options, (6) trap handling improvements, (7) wait -p enhancements, (8) READLINE_ARGUMENT and history expansion changes. Provides: complete 5.3 feature reference, version-detection snippets, compatibility shims for older bash, migration recipes, and POSIX.1-2024 alignment notes.

Critical PowerShell changes, deprecations, and migrations for 2025. PROACTIVELY activate for: (1) PowerShell 7.5 GA changes, (2) PowerShell 7.6 preview features, (3) Windows PowerShell 5.1 deprecation timeline, (4) PSReadLine 2.4+ updates, (5) Az PowerShell 14.x breaking changes, (6) Microsoft.Graph migration from AzureAD/MSOnline modules, (7) PSResourceGet replacing PowerShellGet 2.x, (8) DSC v3 and migration from DSC v1/v2, (9) ExecutionPolicy and constrained language mode updates, (10) .NET 9/.NET 10 integration impact. Provides: deprecation timeline, breaking-change checklist, migration scripts (AzureAD to Microsoft.Graph), and PSResourceGet upgrade steps.

Complete Python package management system. PROACTIVELY activate for: (1) uv package manager (10-100x faster), (2) pyproject.toml configuration, (3) Virtual environment setup, (4) Dependency management with uv.lock, (5) Ruff linting and formatting, (6) src layout project structure, (7) Publishing to PyPI, (8) pip and requirements.txt. Provides: uv commands, pyproject.toml templates, ruff config, pre-commit setup. Ensures modern Python project setup with fast tooling.

Complete Python 3.13+ fundamentals system. PROACTIVELY activate for: (1) Python 3.13 free-threading (no-GIL), (2) JIT compiler usage, (3) Pattern matching syntax, (4) Walrus operator, (5) F-string features 3.12+, (6) Type parameter syntax 3.12+, (7) Exception groups, (8) Dataclasses and enums, (9) Context managers. Provides: Modern syntax, performance features, best practices, naming conventions. Ensures correct Python 3.13+ patterns with optimal performance.

Complete Python gotchas reference. PROACTIVELY activate for: (1) Mutable default arguments, (2) Mutating lists while iterating, (3) is vs == comparison, (4) Late binding in closures, (5) Variable scope (LEGB), (6) Floating point precision, (7) Exception handling pitfalls, (8) Dict mutation during iteration, (9) Circular imports, (10) Class vs instance attributes. Provides: Problem explanations, code examples, fixes for each gotcha. Ensures bug-free Python code.

This skill should be used when the user asks to optimize, export, serve, compress, or accelerate ML inference. PROACTIVELY activate for: (1) latency, throughput, p95/p99, batching, concurrency, KV cache, memory, or cost issues, (2) quantization INT8/INT4, GPTQ, AWQ, bitsandbytes, pruning, sparsity, distillation, (3) ONNX export, ONNX Runtime, TensorRT, TorchScript, torch.compile, XLA, OpenVINO, Core ML, TFLite, (4) Triton, TorchServe, TF Serving, BentoML, Seldon, KServe configuration, (5) edge deployment, CPU/GPU/TPU/Inferentia serving. Provides: hardware-aware inference optimization and safe benchmarking.

Unity performance optimization for slow games, stutters, and FPS drops. PROACTIVELY activate for: (1) game is slow, stuttering, or has FPS drops, (2) Unity Profiler analysis, (3) Frame Debugger and draw-call investigation, (4) batching (static, dynamic, GPU instancing, SRP Batcher), (5) object pooling to reduce GC pressure, (6) garbage collection spikes and allocation hot paths, (7) memory leaks and Memory Profiler usage, (8) LOD and occlusion culling, (9) texture compression and mipmap settings, (10) build size reduction. Provides: profiling workflow, batching patterns, object-pool implementations, GC reduction techniques, LOD/culling setup, and texture-compression matrix.

Kaggle datasets, models, sources, and kagglehub workflows. PROACTIVELY activate for: (1) downloading datasets with kagglehub, (2) uploading datasets, (3) downloading or uploading Kaggle models, (4) competition_download, (5) notebook_output_download, (6) choosing Kaggle CLI vs kagglehub, (7) attaching dataset_sources, competition_sources, kernel_sources, or model_sources, (8) model artifact transfer, (9) source dependency cleanup, (10) kagglehub limitations for notebooks. Provides: dataset/model transfer patterns, source attachment guidance, tool selection, and limitation checks.

Access comprehensive LaTeX templates, formatting requirements, and submission guidelines for major scientific publication venues (Nature, Science, PLOS, IEEE, ACM), academic conferences (NeurIPS, ICML, CVPR, CHI), research posters, and grant proposals (NSF, NIH, DOE, DARPA). This skill should be used when preparing manuscripts for journal submission, conference papers, research posters, or grant proposals and need venue-specific formatting requirements and templates.

Bayesian modeling with PyMC. Build hierarchical models, MCMC (NUTS), variational inference, LOO/WAIC comparison, posterior checks, for probabilistic programming and inference.

Executive leadership guidance for strategic decision-making, organizational development, and stakeholder management. Includes strategy analyzer, financial scenario modeling, board governance frameworks, and investor relations playbooks. Use when planning strategy, preparing board presentations, managing investors, developing organizational culture, making executive decisions, or when user mentions CEO, strategic planning, board meetings, investor updates, organizational leadership, or executive strategy.

# Motion Animation Library

Deep learning framework (PyTorch Lightning). Organize PyTorch code into LightningModules, configure Trainers for multi-GPU/TPU, implement data pipelines, callbacks, logging (W&B, TensorBoard), distributed training (DDP, FSDP, DeepSpeed), for scalable neural network training.

Single-cell RNA-seq analysis. Load .h5ad/10X data, QC, normalization, PCA/UMAP/t-SNE, Leiden clustering, marker genes, cell type annotation, trajectory, for scRNA-seq analysis.

Machine learning in Python with scikit-learn. Use when working with supervised learning (classification, regression), unsupervised learning (clustering, dimensionality reduction), model evaluation, hyperparameter tuning, preprocessing, or building ML pipelines. Provides comprehensive reference documentation for algorithms, preprocessing techniques, pipelines, and best practices.

Process-based discrete-event simulation framework in Python. Use this skill when building simulations of systems with processes, queues, resources, and time-based events such as manufacturing systems, service operations, network traffic, logistics, or any system where entities interact with shared resources over time.

Cheminformatics toolkit for fine-grained molecular control. SMILES/SDF parsing, descriptors (MW, LogP, TPSA), fingerprints, substructure search, 2D/3D generation, similarity, reactions. For standard workflows with simpler interface, use datamol (wrapper around RDKit). Use rdkit for advanced control, custom sanitization, specialized algorithms.

CLI/Python toolkit for rapid bioinformatics queries. Preferred for quick BLAST searches. Access to 20+ databases: gene info (Ensembl/UniProt), AlphaFold, ARCHS4, Enrichr, OpenTargets, COSMIC, genome downloads. For advanced BLAST/batch processing, use biopython. For multi-database integration, use bioservices.

Access USPTO APIs for patent/trademark searches, examination history (PEDS), assignments, citations, office actions, TSDR, for IP analysis and prior art searches.

Therapeutics Data Commons. AI-ready drug discovery datasets (ADME, toxicity, DTI), benchmarks, scaffold splits, molecular oracles, for therapeutic ML and pharmacological prediction.

Senior Regulatory Affairs Manager expertise for HealthTech and MedTech companies. Provides strategic regulatory guidance, submission management, regulatory pathway analysis, global compliance coordination, and cross-functional team leadership. Use for regulatory strategy development, submission planning, regulatory risk assessment, and team coordination activities.

Molecular machine learning toolkit. Property prediction (ADMET, toxicity), GNNs (GCN, MPNN), MoleculeNet benchmarks, pretrained models, featurization, for drug discovery ML.

Query Open Targets Platform for target-disease associations, drug target discovery, tractability/safety data, genetics/omics evidence, known drugs, for therapeutic target identification.

Look up current research information using Perplexity's Sonar Pro or Sonar Reasoning Pro models through OpenRouter. Automatically selects the best model based on query complexity. Search academic papers, recent studies, technical documentation, and general research information with citations.

Efficient database search tool for bioRxiv preprint server. Use this skill when searching for life sciences preprints by keywords, authors, date ranges, or categories, retrieving paper metadata, downloading PDFs, or conducting literature reviews.

Medicinal chemistry filters. Apply drug-likeness rules (Lipinski, Veber), PAINS filters, structural alerts, complexity metrics, for compound prioritization and library filtering.

Comprehensive biosignal processing toolkit for analyzing physiological data including ECG, EEG, EDA, RSP, PPG, EMG, and EOG signals. Use this skill when processing cardiovascular signals, brain activity, electrodermal responses, respiratory patterns, muscle activity, or eye movements. Applicable for heart rate variability analysis, event-related potentials, complexity measures, autonomic nervous system assessment, psychophysiology research, and multi-modal physiological signal integration.

Differential gene expression analysis (Python DESeq2). Identify DE genes from bulk RNA-seq counts, Wald tests, FDR correction, volcano/MA plots, for RNA-seq analysis.

Access ClinPGx pharmacogenomics data (successor to PharmGKB). Query gene-drug interactions, CPIC guidelines, allele functions, for precision medicine and genotype-guided dosing decisions.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

This skill should be used for time series machine learning tasks including classification, regression, clustering, forecasting, anomaly detection, segmentation, and similarity search. Use when working with temporal data, sequential patterns, or time-indexed observations requiring specialized algorithms beyond standard ML approaches. Particularly suited for univariate and multivariate time series analysis with scikit-learn compatible APIs.

Infer gene regulatory networks (GRNs) from gene expression data using scalable algorithms (GRNBoost2, GENIE3). Use when analyzing transcriptomics data (bulk RNA-seq, single-cell RNA-seq) to identify transcription factor-target gene relationships and regulatory interactions. Supports distributed computation for large-scale datasets.

Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam, sparse attention

Senior GDPR/DSGVO expert and internal/external auditor for data protection compliance. Provides EU GDPR and German DSGVO expertise, privacy impact assessments, data protection auditing, and compliance verification. Use for GDPR compliance assessments, privacy audits, data protection planning, and regulatory compliance verification.

Comprehensive Python library for astronomy and astrophysics. This skill should be used when working with astronomical data including celestial coordinates, physical units, FITS files, cosmological calculations, time systems, tables, world coordinate systems (WCS), and astronomical data analysis. Use when tasks involve coordinate transformations, unit conversions, FITS file manipulation, cosmological distance calculations, time scale conversions, or astronomical data processing.

This skill should be used when working with single-cell omics data analysis using scvi-tools, including scRNA-seq, scATAC-seq, CITE-seq, spatial transcriptomics, and other single-cell modalities. Use this skill for probabilistic modeling, batch correction, dimensionality reduction, differential expression, cell type annotation, multimodal integration, and spatial analysis tasks.

Query ChEMBL's bioactive molecules and drug discovery data. Search compounds by structure/properties, retrieve bioactivity data (IC50, Ki), find inhibitors, perform SAR studies, for medicinal chemistry.

Access COSMIC cancer mutation database. Query somatic mutations, Cancer Gene Census, mutational signatures, gene fusions, for cancer research and precision oncology. Requires authentication.

Senior FDA consultant and specialist for medical device companies including HIPAA compliance and requirement management. Provides FDA pathway expertise, QSR compliance, cybersecurity guidance, and regulatory submission support. Use for FDA submission planning, QSR compliance assessments, HIPAA evaluations, and FDA regulatory strategy development.

Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean model implementations, educational understanding of architectures, or production fine-tuning with LoRA/QLoRA. Single-file implementations, no abstraction layers.

Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limited compute (5× cost reduction vs dense models), implementing sparse architectures like Mixtral 8x7B or DeepSeek-V3, or scaling model capacity without proportional compute increase. Covers MoE architectures, routing mechanisms, load balancing, expert parallelism, and inference optimization.

Comprehensive healthcare AI toolkit for developing, testing, and deploying machine learning models with clinical data. This skill should be used when working with electronic health records (EHR), clinical prediction tasks (mortality, readmission, drug recommendation), medical coding systems (ICD, NDC, ATC), physiological signals (EEG, ECG), healthcare datasets (MIMIC-III/IV, eICU, OMOP), or implementing deep learning models for healthcare applications (RETAIN, SafeDrug, Transformer, GNN).

Laboratory automation toolkit for controlling liquid handlers, plate readers, pumps, heater shakers, incubators, centrifuges, and analytical equipment. Use this skill when automating laboratory workflows, programming liquid handling robots (Hamilton STAR, Opentrons OT-2, Tecan EVO), integrating lab equipment, managing deck layouts and resources (plates, tips, containers), reading plates, or creating reproducible laboratory protocols. Applicable for both simulated protocols and physical hardware control.

Use this skill when working with symbolic mathematics in Python. This skill should be used for symbolic computation tasks including solving equations algebraically, performing calculus operations (derivatives, integrals, limits), manipulating algebraic expressions, working with matrices symbolically, physics calculations, number theory problems, geometry computations, and generating executable code from mathematical expressions. Apply this skill when the user needs exact symbolic results rather than numerical approximations, or when working with mathematical formulas that contain variables and parameters.

Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Use when training models >1B parameters, need maximum GPU efficiency (47% MFU on H100), or require tensor/pipeline/sequence/context/expert parallelism. Production-ready framework used for Nemotron, LLaMA, DeepSeek.

Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support

EU MDR 2017/745 regulation specialist and consultant for medical device requirement management. Provides comprehensive MDR compliance expertise, gap analysis, technical documentation guidance, clinical evidence requirements, and post-market surveillance implementation. Use for MDR compliance assessment, classification decisions, technical file preparation, and regulatory requirement interpretation.

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

Computational pathology toolkit for analyzing whole-slide images (WSI) and multiparametric imaging data. Use this skill when working with histopathology slides, H&E stained images, multiplex immunofluorescence (CODEX, Vectra), spatial proteomics, nucleus detection/segmentation, tissue graph construction, or training ML models on pathology data. Supports 160+ slide formats including Aperio SVS, NDPI, DICOM, OME-TIFF for digital pathology workflows.

Senior Quality Manager Responsible Person (QMR) for HealthTech and MedTech companies. Provides overall quality system responsibility, regulatory compliance oversight, management accountability, and strategic quality leadership. Use for quality system governance, regulatory compliance oversight, management responsibility, and quality strategic planning.

Comprehensive toolkit for survival analysis and time-to-event modeling in Python using scikit-survival. Use this skill when working with censored survival data, performing time-to-event analysis, fitting Cox models, Random Survival Forests, Gradient Boosting models, or Survival SVMs, evaluating survival predictions with concordance index or Brier score, handling competing risks, or implementing any survival analysis workflow with the scikit-survival library.

Graph-based drug discovery toolkit. Molecular property prediction (ADMET), protein modeling, knowledge graph reasoning, molecular generation, retrosynthesis, GNNs (GIN, GAT, SchNet), 40+ datasets, for PyTorch-based ML on molecules, proteins, and biomedical graphs.

Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support

Senior CAPA Officer specialist for managing Corrective and Preventive Actions within Quality Management Systems. Provides CAPA process management, root cause analysis, effectiveness verification, and continuous improvement coordination. Use for CAPA investigations, corrective action planning, preventive action implementation, and CAPA system optimization.

Framework for computational fluid dynamics simulations using Python. Use when running fluid dynamics simulations including Navier-Stokes equations (2D/3D), shallow water equations, stratified flows, or when analyzing turbulence, vortex dynamics, or geophysical flows. Provides pseudospectral methods with FFT, HPC support, and comprehensive output analysis.

Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learning transformers. By Andrej Karpathy. Perfect for understanding GPT architecture from scratch. Train on Shakespeare (CPU) or OpenWebText (multi-GPU).

Comprehensive toolkit for creating, analyzing, and visualizing complex networks and graphs in Python. Use when working with network/graph data structures, analyzing relationships between entities, computing graph algorithms (shortest paths, centrality, clustering), detecting communities, generating synthetic networks, or visualizing network topologies. Applicable to social networks, biological networks, transportation systems, citation networks, and any domain involving pairwise relationships.

Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local models (Transformers, vLLM), and maximize inference speed with Outlines - dottxt.ai's structured generation library