Claude Skills by thedixitjain

Use when an accepted SIGIR paper must become the ACM Digital Library version of record — de-anonymizing the sigconf source, CCS concepts and keywords, the e-rights form, TAPS source upload and proof checking, restoring real system and dataset names, and the April camera-ready window before the July conference.

Use when packaging an ACM SIGMETRICS artifact for the ACM Artifact Review and Badging scheme (Artifacts Available, Evaluated Functional and Reusable, Results Reproduced), covering what performance-evaluation evaluators check first (does the simulation regenerate the figures and match the analysis?), DOI-issuing archives, evaluator-proof documentation, and confirming whether an artifact track runs this cycle.

Use when preparing a SIGMOD paper's code and data for the Availability & Reproducibility Initiative (ARI), covering the post-acceptance opt-in, HotCRP artifact registration, the Artifacts Available / Artifacts Evaluated / Results Reproduced badges, evaluator criteria, and the Best Artifact award path.

Use when designing or auditing the evaluation of a SIGMOD paper, covering workload realism and standard benchmark usage, baseline tuning fairness, scalability and tail-latency methodology, ablations that isolate the mechanism, and the setup disclosure a data-systems PC demands before trusting any speedup.

Reference-free preference alignment, simpler than DPO.

RL post-training for LLMs with Megatron and SGLang.

Use when packaging an ACM SoCC artifact for the ACM Artifact Review and Badging scheme (Artifacts Available, Evaluated Functional and Reusable, Results Reproduced), covering what a cloud-systems evaluator checks first, reproducing tail-latency and cost results on a testbed, DOI-issuing archives, and the fact that whether SoCC runs a dedicated artifact-evaluation track for a given edition must be verified.

Use when deciding how code, computations, or data attach to a SODA (ACM-SIAM Symposium on Discrete Algorithms) paper — SODA itself runs no artifact track, so this skill covers computer-assisted proof evidence, implementation companions, and when to route the artifact to co-located ALENEX's formal artifact evaluation instead.

Activate for ANY startup evaluation, investment screening, or company assessment. Triggers include: \"evaluate this startup\", \"screen this company\", \"should I invest in X\", \"is this a good investment\", \"what do you think about this company\", \"review this startup\", \"score this company\", \"rate this pitch\", \"assess this founder\", \"quick take on X\", \"is X worth investing in\", \"pass or decline on X\", \"what's your verdict on X\", \"first look at this company\", \"quick scree...

Use when preparing a SOSP artifact for the post-acceptance evaluation run by the sysartifacts community, registering within days of notification, packaging for the cooperative review process, and targeting the ACM badges — Artifacts Available, Artifacts Evaluated Functional, and Artifacts Evaluated Reusable.

Use when designing or auditing the evaluation of a SOSP paper — mapping every claim to an experiment, choosing baselines a systems PC will accept as fair, mixing microbenchmarks with end-to-end and failure runs, reporting tails and overheads honestly, and isolating the mechanism the design credits.

Use when planning a SOSP campaign across the annual cycle — back-scheduling from the spring abstract and paper deadlines, running the summer response and July notification, stacking the August camera-ready with artifact evaluation, and retargeting rejections across the SOSP/OSDI/EuroSys circuit.

Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.

Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.

Train sparse autoencoders to interpret model features.

'Execute Speak secondary workflow: Pronunciation Training with phoneme-level analysis. Use when implementing pronunciation drills, speech scoring, or targeted pronunciation improvement features. Trigger with phrases like \"speak pronunciation training\", \"speak speech scoring\", \"speak phoneme analysis\". '

当为《中国科学：信息科学》(Scientia Sinica Informationis, SSI) 稿件准备代码与数据可用性材料、需要判断本刊是否设有独立\"制品评审/Artifact Evaluation\"环节及如何据实处理时调用。本刊作为中文信息科学综合旗舰，并未采用英文会议式的独立制品徽章流程（现状：待核实/以官网当期为准），本技能说明如何在没有强制徽章制度的前提下，主动提供可用、可查、可复现的代码与数据，撰写诚实的可用性声明，并与英文姊妹刊 Science China Information Sciences 的开放科学要求对照。适用于计算机、控制、通信、微电子等各子学科稿件。

Production-ready reinforcement learning algorithms (PPO, SAC, DQN, TD3, DDPG, A2C) with scikit-learn-like API. Use for standard RL experiments, quick prototyping, and well-documented algorithm implementations. Best for single-agent RL with Gymnasium environments. For high-performance parallel training, multi-agent systems, or custom vectorized environments, use pufferlib instead.

Text-to-image generation, inpainting, and img2img.

Guided statistical analysis for research data - test selection, assumption checking, effect sizes, power analysis, Bayesian alternatives, and APA-formatted reporting. Use whenever a user wants to compare groups, test a hypothesis, analyze experimental or survey data, check statistical assumptions, compute required sample sizes, or write up results - even if they never name a specific test. Covers t-tests, ANOVA, chi-square, correlation, regression, non-parametric and Bayesian methods. For low...

Run hypothesis tests, analyze A/B experiment results, calculate sample sizes, and interpret statistical significance with effect sizes. Use when you need to validate whether observed differences are real, size an experiment correctly before launch, or interpret test results with confidence.

../../../engineering/statistical-analyst/skills/statistical-analyst/SKILL.md

Sample-size and statistical power calculations for planning studies. Use whenever someone asks \"how many subjects/samples/replicates do I need\", wants an a priori power analysis, a minimum detectable effect (MDE), a power curve, or needs to justify a sample size for a grant, IRB protocol, or pre-registration. Covers closed-form power for t-tests, ANOVA, proportions, correlations, chi-square, and regression, plus simulation-based (Monte Carlo) power for designs with no formula — logistic/Poi...

'Configure and manage - Calculate statistical significance calculator operations. Auto-activating skill for Data Analytics. Triggers on: statistical significance calculator, statistical significance calculator Part of the Data Analytics skill category. Use when working with statistical significance calculator functionality. Trigger with phrases like \"statistical significance calculator\", \"statistical calculator\", \"statistical\". '

Statistical models library for Python. Use when you need specific model classes (OLS, GLM, mixed models, ARIMA) with detailed diagnostics, residuals, and inference. Best for econometrics, time series, rigorous inference with coefficient tables. For guided statistical test selection with APA reporting use statistical-analysis.

Statsmodels is Python's premier library for statistical modeling, providing tools for estimation, inference, and diagnostics across a wide range of statistical methods.

Use when deciding what evidence objects a STOC (ACM Symposium on Theory of Computing) paper must ship, given that STOC runs no artifact-evaluation track — the durable artifact is the public full version on arXiv/ECCC, plus verifiable certificates whenever a proof leans on computation, and optional mechanization.

Use when judging whether computation belongs in a STOC (ACM Symposium on Theory of Computing) paper at all — STOC accepts on theorems, with no empirical-evaluation expectation — and when it does, scoping it as certified proof computation, constructive search, or clearly labeled illustration rather than benchmarking.

'Configure streaming inference setup operations. Auto-activating skill for ML Deployment. Triggers on: streaming inference setup, streaming inference setup Part of the ML Deployment skill category. Use when working with streaming inference setup functionality. Trigger with phrases like \"streaming inference setup\", \"streaming setup\", \"streaming\". '

Growth experiment design — structure a growth hypothesis, define metric, baseline, expected lift, and kill condition for a single experiment. Use when asked to \"design a growth experiment\", \"test this growth idea\", \"experiment framework\", \"how do we test if this works\", or \"growth hypothesis\".

Use when you need exact symbolic math in Python — algebra, calculus, equation solving, symbolic linear algebra, or code generation via lambdify/LaTeX. Prefer NumPy or SciPy when floating-point approximations are sufficient.

Use when packaging a TACAS (ETAPS) artifact for the ETAPS Artifact Badges (Available, Functional, Reusable), covering the two-round model (mandatory, PC-parallel evaluation for tool and tool-demo papers vs voluntary post-acceptance evaluation for research and case-study papers), what the AEC checks on the clean evaluation VM, DOI-issuing archives, and evaluator-proof documentation.

Use when drafting a TACAS (ETAPS) author response for the rebuttal window, covering the short single-round rebuttal, answering soundness and benchmark-fairness objections with checkable evidence, addressing artifact concerns for tool papers, respecting the per-category blind model when it applies, and reading which objections a rebuttal can and cannot move.

'Generate tensorboard visualizer operations. Auto-activating skill for ML Training. Triggers on: tensorboard visualizer, tensorboard visualizer Part of the ML Training skill category. Use when working with tensorboard visualizer functionality. Trigger with phrases like \"tensorboard visualizer\", \"tensorboard visualizer\", \"tensorboard\". '

'Build tensorflow model trainer operations. Auto-activating skill for ML Training. Triggers on: tensorflow model trainer, tensorflow model trainer Part of the ML Training skill category. Use when working with tensorflow model trainer functionality. Trigger with phrases like \"tensorflow model trainer\", \"tensorflow trainer\", \"tensorflow\". '

'Create tensorflow savedmodel creator operations. Auto-activating skill for ML Deployment. Triggers on: tensorflow savedmodel creator, tensorflow savedmodel creator Part of the ML Deployment skill category. Use when working with tensorflow savedmodel creator functionality. Trigger with phrases like \"tensorflow savedmodel creator\", \"tensorflow creator\", \"tensorflow\". '

'Configure tensorflow serving setup operations. Auto-activating skill for ML Deployment. Triggers on: tensorflow serving setup, tensorflow serving setup Part of the ML Deployment skill category. Use when working with tensorflow serving setup functionality. Trigger with phrases like \"tensorflow serving setup\", \"tensorflow setup\", \"tensorflow\". '

High-throughput LLM inference on NVIDIA GPUs.

'Manage time series decomposer operations. Auto-activating skill for Data Analytics. Triggers on: time series decomposer, time series decomposer Part of the Data Analytics skill category. Use when working with time series decomposer functionality. Trigger with phrases like \"time series decomposer\", \"time decomposer\", \"time\". '

Zero-shot time series forecasting with Google's TimesFM foundation model. Use for any univariate time series (sales, sensors, energy, vitals, weather) without training a custom model. Supports CSV/DataFrame/array inputs with point forecasts and prediction intervals. Includes a preflight system checker script to verify RAM/GPU before first use.

'Together AI common errors for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together common errors\". '

'Together AI core workflow a for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together core workflow a\". '

'Together AI core workflow b for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together core workflow b\". '

'Together AI cost tuning for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together cost tuning\". '

'Install Together AI SDK and configure API key for inference and fine-tuning. Use when setting up Together AI, configuring the OpenAI-compatible API, or initializing the together Python package. Trigger: \"install together, setup together ai, together API key\". '

'Together AI local dev loop for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together local dev loop\". '

'Together AI rate limits for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together rate limits\". '

'Together AI sdk patterns for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together sdk patterns\". '

'Together AI webhooks events for inference, fine-tuning, and model deployment. Use when working with Together AI''s OpenAI-compatible API. Trigger: \"together webhooks events\". '

PyTorch Geometric (PyG) for graph neural networks — node/link/graph classification, message passing (GCN, GAT, GraphSAGE, GIN), heterogeneous graphs, neighbor sampling, and custom datasets. Use when working with torch_geometric, not for general NetworkX analytics or non-graph PyTorch models.