Skill collection · Orchestra-Research
AI-Research-SKILLs agent skills by Orchestra-Research
81 skills from Orchestra-Research/AI-Research-SKILLs are listed in the directory, across 7 categories. 12 skills have a full page with install steps and a pre-install summary.
View Orchestra-Research/AI-Research-SKILLs on GitHub
At a glance
| Publisher | Orchestra-Research |
|---|---|
| Repository | github.com/Orchestra-Research/AI-Research-SKILLs |
| Listed skills | 81 |
| Security scan | Not scanned |
| Categories | AI Agents and LLM Engineering, Science and Research, Data, Analytics and Databases, DevOps and Cloud, Security, Documents and Office Files, Video, Image, Audio and Creative |
| License | MIT |
| Last commit | June 15, 2026 |
| Install every skill | npx skills add https://github.com/Orchestra-Research/AI-Research-SKILLs |
| README badge | Get the badge for this listing |
Show the listing in your README
This general badge links to the repository listing. Each skill keeps its own score and grade; skills scoring 80–99 add a numeric and a letter quality badge. Choose Improve this score beside a skill below to request a fresh review.
Markdown
[](https://skills.suedeai.ai/directory/orchestra-research/ai-research-skills/)HTML
<a href="https://skills.suedeai.ai/directory/orchestra-research/ai-research-skills/"><img src="https://skills.suedeai.ai/directory/orchestra-research/ai-research-skills/badge.svg" alt="Suede Skills Directory: listed" height="20"></a>Limited-time feature offer: display either the numeric or letter badge in your repository or skill README (both also qualify), then request badge verification. Eligible skills scoring 80–99 can receive a featured placement after Suede verifies at least one valid badge linked to the skill’s listing. Suede has not announced an end date.
Skills in Orchestra-Research/AI-Research-SKILLs
Linked names open the skill's directory page; the rest open the skill folder on GitHub. Descriptions on rows without a page are quoted from each SKILL.md.
No matching skills. Try fewer words or clear your search.
| Skill | Category | What it does |
|---|---|---|
| academic-plotting 77/100 · C | Science | Generates publication-quality figures for ML papers, including architecture diagrams via Gemini and data-driven charts via matplotlib or seaborn, from research context. |
| faiss 77/100 · C | Data | Guides fast k-NN search and clustering of dense vectors with Facebook's FAISS, including GPU acceleration and Flat, IVF, and HNSW index types. |
| knowledge-distillation 77/100 · C | AI agents | Compresses large language models into smaller student models using knowledge distillation, covering temperature scaling, soft targets, reverse KLD, logit distillation, and MiniLLM training. |
| langchain 77/100 · C | AI agents | Guides building LLM-powered applications with LangChain, including agents, chains, RAG, tool calling, memory management, and vector store retrieval across multiple providers. |
| llava 77/100 · C | AI agents | Guides visual instruction tuning and image-based conversations with LLaVA, which combines a CLIP vision encoder with Vicuna or LLaMA models for visual question answering. |
| ml-training-recipes 77/100 · C | AI agents | Provides PyTorch training recipes for LLMs, vision, diffusion, and genomics, covering training loops, optimizers, learning rate schedules, mixed precision, debugging, and GPU throughput. |
| model-merging 77/100 · C | AI agents | Merges multiple fine-tuned models with mergekit to combine capabilities without retraining, using methods such as SLERP, TIES-Merging, DARE, and Task Arithmetic. |
| model-pruning 77/100 · C | AI agents | Reduces LLM size and speeds inference with pruning techniques like Wanda and SparseGPT, without retraining, including structured, unstructured, N:M, and magnitude pruning. |
| sentence-transformers 77/100 · C | AI agents | Generates sentence, text, and image embeddings with 5000+ pre-trained models for semantic similarity, clustering, retrieval, RAG, and semantic search. |
| sentencepiece 77/100 · C | AI agents | Guides training language-independent SentencePiece tokenizers with BPE or Unigram directly on raw text, for multilingual, CJK, and reproducible tokenization needs. |
| speculative-decoding 77/100 · C | AI agents | Accelerates LLM inference with speculative decoding, Medusa multiple heads, and lookahead decoding, covering draft models, tree-based attention, and parallel token generation. |
| systems-paper-writing 77/100 · C | Science | Guides writing systems papers for OSDI, SOSP, ASPLOS, NSDI, and EuroSys with structural blueprints, writing patterns, venue checklists, reviewer guidelines, LaTeX templates, and deadlines. |
| ara-compiler 69/100 · D | Science | Compiles any research input — PDF papers, GitHub repositories, experiment logs, code directories, or raw notes — into a complete Agent-Native… |
| autogpt-agents 69/100 · D | AI agents | Autonomous AI agent platform for building and deploying continuous agents. Use when creating visual workflow agents, deploying persistent autonomous… |
| autoresearch 66/100 · D | Science | Orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid experiment iterations with clear… |
| awq-quantization 69/100 · D | AI agents | Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when deploying large models (7B-70B)… |
| axolotl 67/100 · D | AI agents | Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support |
| brainstorming-research-ideas 74/100 · C | Science | Guides researchers through structured ideation frameworks to discover high-impact research directions. Use when exploring new problem spaces… |
| chroma 77/100 · C | Data | Open-source embedding database for AI applications. Store embeddings and metadata, perform vector and full-text search, filter by metadata. Simple… |
| clip 77/100 · C | AI agents | OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross-modal retrieval. Trained on… |
| constitutional-ai 74/100 · C | AI agents | Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with self-critique/revision, then… |
| creative-thinking-for-research 71/100 · C | Science | Applies cognitive science frameworks for creative thinking to CS and AI research ideation. Use when seeking genuinely novel research directions by… |
| crewai-multi-agent 69/100 · D | AI agents | Multi-agent orchestration framework for autonomous AI collaboration. Use when building teams of specialized agents working together on complex… |
| deepspeed 61/100 · D | AI agents | Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam, sparse attention |
| distributed-llm-pretraining-torchtitan 69/100 · D | AI agents | Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1… |
| evaluating-code-models 69/100 · D | AI agents | Evaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when benchmarking code models… |
| evaluating-cosmos-policy 69/100 · D | AI agents | Evaluates NVIDIA Cosmos Policy on LIBERO and RoboCasa simulation environments. Use when setting up cosmos-policy for robot manipulation evaluation… |
| evaluating-llms-harness 69/100 · D | AI agents | Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing… |
| evolving-ai-agents 69/100 · D | AI agents | Provides guidance for automatically evolving and optimizing AI agents across any domain using LLM-driven evolution algorithms. Use when building… |
| experiment-tracking-swanlab 69/100 · D | AI agents | Provides guidance for experiment tracking with SwanLab. Use when you need open-source run tracking, local or self-hosted dashboards, and lightweight… |
| fine-tuning-openvla-oft 69/100 · D | AI agents | Fine-tunes and evaluates OpenVLA-OFT and OpenVLA-OFT+ policies for robot action generation with continuous action heads, LoRA adaptation, and FiLM… |
| fine-tuning-serving-openpi 69/100 · D | AI agents | Fine-tune and serve Physical Intelligence OpenPI models (pi0, pi0-fast, pi0.5) using JAX or PyTorch backends for robot policy inference across… |
| fine-tuning-with-trl 69/100 · D | AI agents | Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization… |
| gguf-quantization 69/100 · D | AI agents | GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when… |
| gptq 77/100 · C | AI agents | Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4×… |
| hqq-quantization 69/100 · D | AI agents | Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration… |
| huggingface-tokenizers 69/100 · D | AI agents | Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Supports BPE, WordPiece, and Unigram… |
| implementing-llms-litgpt 69/100 · D | AI agents | Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean… |
| langsmith-observability 69/100 · D | AI agents | LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets… |
| llama-cpp 77/100 · C | AI agents | Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M3 Macs, AMD/Intel GPUs, or when… |
| llama-factory 64/100 · D | AI agents | Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support |
| llamaguard 66/100 · D | AI agents | Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual content, weapons, substances… |
| llamaindex 69/100 · D | AI agents | Data framework for building LLM applications with RAG. Specializes in document ingestion (300+ connectors), indexing, and querying. Features vector… |
| long-context 69/100 · D | AI agents | Extend context windows of transformer models using RoPE, YaRN, ALiBi, and position interpolation techniques. Use when processing long documents… |
| miles-rl-training 69/100 · D | AI agents | Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training large MoE models with FP8/INT4… |
| ml-paper-writing 67/100 · D | Science | Write publication-ready ML/AI papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM. Use when drafting papers from research repos, structuring arguments… |
| modal-serverless-gpu 69/100 · D | DevOps | Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure management, deploying ML… |
| moe-training 69/100 · D | AI agents | Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limited compute (5× cost reduction… |
| nanogpt 69/100 · D | AI agents | Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learning transformers. By Andrej… |
| nemo-curator 77/100 · C | AI agents | GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+… |
| nemo-evaluator-sdk 69/100 · D | AI agents | Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend execution. Use when needing… |
| nemo-guardrails 66/100 · D | Security | NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-checking, hallucination… |
| nnsight-remote-interpretability 69/100 · D | AI agents | Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to… |
| openrlhf-training 69/100 · D | AI agents | High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM… |
| optimizing-attention-flash 69/100 · D | AI agents | Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long… |
| peft-fine-tuning 69/100 · D | AI agents | Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory… |
| phoenix-observability 69/100 · D | AI agents | Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM applications with detailed traces, running… |
| pinecone 77/100 · C | Data | Managed vector database for production AI applications. Fully managed, auto-scaling, with hybrid search (dense + sparse), metadata filtering, and… |
| presenting-conference-talks 77/100 · C | Documents | Generates conference presentation slides (Beamer LaTeX PDF and editable PPTX) from a compiled paper with speaker notes and talk script. Use when… |
| prompt-guard 66/100 · D | Security | Meta's 86M prompt injection and jailbreak detector. Filters malicious prompts and third-party data for LLM apps. 99%+ TPR, <1% FPR. Fast (<2ms GPU)… |
| pytorch-fsdp2 77/100 · C | AI agents | Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and distributed checkpointing. Use… |
| pytorch-lightning 77/100 · C | AI agents | High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate… |
| pyvene-interventions 69/100 · D | AI agents | Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention framework. Use when conducting… |
| qdrant-vector-search 69/100 · D | Data | High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest… |
| quantizing-models-bitsandbytes 69/100 · D | AI agents | Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limited, need to fit larger models… |
| ray-data 77/100 · C | Data | Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Integrates with Ray Train, PyTorch… |
| ray-train 77/100 · C | AI agents | Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter… |
| segment-anything-model 69/100 · D | AI agents | Foundation model for image segmentation with zero-shot transfer. Use when you need to segment any object in images using points, boxes, or masks as… |
| serving-llms-vllm 69/100 · D | AI agents | Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production LLM APIs, optimizing inference… |
| sglang 77/100 · C | AI agents | Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, constrained decoding, agentic… |
| simpo-training 69/100 · D | AI agents | Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No… |
| skypilot-multi-cloud-orchestration 69/100 · D | DevOps | Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds… |
| slime-rl-training 69/100 · D | AI agents | Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM models, implementing custom data… |
| sparse-autoencoder-training 69/100 · D | AI agents | Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable… |
| tensorrt-llm 77/100 · C | AI agents | Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100)… |
| torchforge-rl-training 69/100 · D | AI agents | Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Use when you want clean RL… |
| training-llms-megatron 69/100 · D | AI agents | Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Use when training models >1B… |
| transformer-lens-interpretability 69/100 · D | AI agents | Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and… |
| unsloth 64/100 · D | AI agents | Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization |
| verl-rl-training 69/100 · D | AI agents | Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL… |
| whisper 77/100 · C | Media | OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six… |
Where this repository is cited
Orchestra-Research/AI-Research-SKILLs appears in 2 curated lists: VoltAgent/awesome-agent-skills, skillmatic-ai/awesome-agent-skills.