guidance
97Skill Claude CodeCodex
Constrain LLM output with grammars; guarantee valid JSON.
not rated 15
+1 28d ago A 16 tokens
copy · 100% MIT
15Stars on the repository
191Mods indexed here, across every type
28d agoLast push, which is what freshness is scored on
MITLicence, which decides whether bodies are shown
Skill Claude CodeCodex
HuggingFace hf CLI: search/download/upload models, datasets.
not rated 15
+1 28d ago C 20 tokens
copy · 100% MIT
Skill Claude CodeCodex
Fast BPE/WordPiece tokenization and custom vocab training.
not rated 15
+1 28d ago A 19 tokens
copy · 100% MIT
llama-cpp
100Skill Claude CodeCodex
Use this skill for local GGUF inference, quant selection, or Hugging Face repo discovery for llama.cpp.
not rated 15
+1 28d ago A 18 tokens
copy · 100% MIT
outlines
101Skill Claude CodeCodex
Outlines: structured JSON/regex/Pydantic LLM generation.
not rated 15
+1 28d ago A 17 tokens
copy · 100% MIT
Skill Claude CodeCodex
Use when deploying production LLM APIs, optimizing inference latency/throughput, or serving models with limited GPU memory. Supports OpenAI-compatible endpoints, quantization (GPTQ/AWQ/FP8), and tensor parallelism.
not rated 15
+1 28d ago A 27 tokens
copy · 100% MIT
instructor
103Skill Claude CodeCodex
Structured LLM outputs validated with Pydantic.
not rated 15
+1 28d ago A 13 tokens
copy · 100% MIT
Skill Claude CodeCodex
On-demand GPU cloud instances for ML training.
not rated 15
+1 28d ago C 16 tokens
copy · 100% MIT
llava
105Skill Claude CodeCodex needs its repo
Vision-language chat: VQA, captioning, image dialogue.
not rated 15
+1 28d ago A 16 tokens
copy · 100% MIT
Skill Claude CodeCodex
Serverless GPU cloud for ML jobs and model APIs.
not rated 15
+1 28d ago A 17 tokens
copy · 91% MIT
Skill Claude CodeCodex needs its repo
SAM: zero-shot image segmentation via points, boxes, masks.
not rated 15
+1 28d ago C 18 tokens
copy · 100% MIT
nemo-curator
108Skill Claude CodeCodex
Curate LLM training data: dedupe, filter, PII redaction.
not rated 15
+1 28d ago A 22 tokens
copy · 100% MIT
obliteratus
109Skill Claude CodeCodex needs its repo
OBLITERATUS: abliterate LLM refusals (diff-in-means).
not rated 15
+1 28d ago A 21 tokens
copy · 100% MIT
peft-fine-tuning
110Skill Claude CodeCodex
Fine-tune large LLMs with LoRA on limited GPU memory.
not rated 15
+1 28d ago A 22 tokens
copy · 94% MIT
pinecone
111Skill Claude CodeCodex
Managed vector DB for production RAG and search.
not rated 15
+1 28d ago A 13 tokens
copy · 100% MIT
pytorch-fsdp
112Skill Claude CodeCodex
Fully sharded data-parallel training for large models.
not rated 15
+1 28d ago B 17 tokens
copy · 98% MIT
Skill Claude CodeCodex
Clean training loops with built-in distributed support.
not rated 15
+1 28d ago A 14 tokens
copy · 100% MIT
Skill Claude CodeCodex
Vector search engine for production RAG systems.
not rated 15
+1 28d ago A 15 tokens
copy · 95% MIT
dspy
115Skill Claude CodeCodex
DSPy: declarative LM programs, auto-optimize prompts, RAG.
not rated 15
+1 28d ago A 19 tokens
copy · 100% MIT
Skill Claude CodeCodex
Train sparse autoencoders to interpret model features.
not rated 15
+1 28d ago A 16 tokens
copy · 97% MIT
simpo-training
117Skill Claude CodeCodex needs its repo
Reference-free preference alignment, simpler than DPO.
not rated 15
+1 28d ago A 14 tokens
copy · 98% MIT
Skill Claude CodeCodex
RL post-training for LLMs with Megatron and SGLang.
not rated 15
+1 28d ago A 21 tokens
copy · 97% MIT
Skill Claude CodeCodex
Text-to-image generation, inpainting, and img2img.
not rated 15
+1 28d ago A 20 tokens
copy · 95% MIT
tensorrt-llm
120Skill Claude CodeCodex
High-throughput LLM inference on NVIDIA GPUs.
not rated 15
+1 28d ago A 18 tokens
copy · 95% MIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: