Every Hermes skill, in one place

185 official skills from the Hermes Agent repository plus 101,376 community entries. Community index checked 2026-11-20 (UTC). Read community skills at their original sources.

Open baoyu-article-illustrator in the official library

Community index generated 2026-11-20 (UTC), checked 2026-11-20 (UTC).

Skill results

34 official + 4,295 community matches. Page 161 of 180.

Official skills first, then community.

Best match / quality.

Microbial population dynamics modeling and analysis. Bacterial growth curve fitting (logistic, Gompertz, Baranyi), Lotka-Volterra community dynamics, Gillespie stochastic simulation, biofilm quantification, CFU enumeration, and genome annot

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/biology/microbial-dynamics
AI models & providersSource (opens in a new tab)

Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training large MoE models with FP8/INT4, needing train-inference alignment, or requiring speculative RL for maximum throughput.

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-inference/miles
AI models & providersSource (opens in a new tab)

Rigorous methodology for evaluating ML models on established benchmarks. Covers proper train/val/test splits, baseline verification from original papers, exact metric formula discrepancies, data-leak detection checklist, multi-seed robustne

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/ml-benchmark-evaluation
Coding & developer tools · AI models & providersSource (opens in a new tab)

Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/mlflow
Coding & developer tools · AI models & providersSource (opens in a new tab)

Modal is a serverless cloud platform for running Python on demand, including on-demand GPUs. Use when deploying or serving AI/ML models, running GPU-accelerated workloads (training, fine-tuning, inference), serving web endpoints, scheduling

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install K-Dense-AI/scientific-agent-skills/skills/modal
AI models & providers · Web, research & browsingSource (opens in a new tab)

Cost modeling and ROI analysis for specialized LLM development. Use when deciding whether to train a custom model, estimating total cost, or calculating break-even vs frontier APIs. Covers training costs, inference costs, and time-to-ROI pr

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/model-economics
AI models & providers · Cost, speed & monitoringSource (opens in a new tab)

Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or ex

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/model-merging
AI models & providers · Chat apps & voiceSource (opens in a new tab)

Implement an upstream pretrained model natively in BioIR using its optimized components, with checkpoint conversion and numerical validation. Use when adding a new model or architecture to BioIR; includes data pipeline integration when requ

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install clawhub/bioir-model-onboard
AI models & providersSource (opens in a new tab)

Implement an upstream pretrained model natively in BioIR using its optimized components, with checkpoint conversion and numerical validation. Use when adding a new model or architecture to BioIR; includes data pipeline integration when requ

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install NVIDIA/skills/skills/bioir-model-onboard
AI models & providersSource (opens in a new tab)

Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT. Use when compressing models without retraining, achieving 50% sparsity with minimal accuracy loss, or enabling faster inference on hardware accelera

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/model-pruning
AI models & providersSource (opens in a new tab)

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 scal

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/moe-training
AI models & providers · Cost, speed & monitoringSource (opens in a new tab)

Molecular featurization for ML (100+ featurizers). ECFP, MACCS, descriptors, pretrained models (ChemBERTa), convert SMILES to features, for QSAR and molecular ML.

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/chemistry/molfeat
AI models & providersSource (opens in a new tab)

Generate multiple sequence alignments (MSAs) for protein sequences using the ColabFold MSA-Search NIM. Use for homolog search, UniRef30/ColabFold env searches, A3M or FASTA alignments, paired MSA search for complexes, PDB70 structural templ

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install NVIDIA/skills/skills/bionemo-msa-search-nim
AI models & providers · Web, research & browsingSource (opens in a new tab)

Pareto-aware molecular design balancing multiple ADMET properties simultaneously. Based on MultiMol (Yu 2025) and MOLLM (Ran 2025).

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/coding/multi-objective-optimization
AI models & providers · Images, video, music & 3DSource (opens in a new tab)

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

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/nanogpt
Coding & developer tools · AI models & providersSource (opens in a new tab)

Queries the NCATS Translator ARAX production API for bounded, typed, provenance-rich one-hop and endpoint-pinned two-hop biomedical knowledge-graph relationships. Use for Biolink-constrained RTX-KG2 lookup, explicit selected-provider ARAX f

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install K-Dense-AI/scientific-agent-skills/skills/ncats-arax
Memory & notes · AI models & providersSource (opens in a new tab)

Queries the NCATS Translator ARAX production API for bounded, typed, provenance-rich one-hop and endpoint-pinned two-hop biomedical knowledge-graph relationships. Use for Biolink-constrained RTX-KG2 lookup, explicit selected-provider ARAX f

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/databases/ncats-arax
Memory & notes · AI models & providersSource (opens in a new tab)

Guide for onboarding new model architectures into NeMo AutoModel, including architecture discovery, implementation patterns, registration, and validation.

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install NVIDIA/skills/skills/nemo-automodel-model-onboarding
AI models & providersSource (opens in a new tab)

GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scales across GPUs with R

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/nemo-curator
AI models & providers · Images, video, music & 3DSource (opens in a new tab)

Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend execution. Use when needing scalable evaluation on local Docker, Slurm HPC, or cloud platforms. NVIDIA's enterprise-grade plat

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/ml-training/nemo-evaluator
Coding & developer tools · AI models & providersSource (opens in a new tab)

NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-checking, hallucination detection, PII filtering, toxicity detection. Uses Colang 2.0 DSL for programmable rails. Production

Community skill. Read the source first. Listed does not mean tested with Hermes.

hermes skills install synthetic-sciences/openscience/backend/cli/skills/llm-tools/nemo-guardrails
AI models & providers · Security & privacySource (opens in a new tab)

Entries with unclear source links are withheld. Some restricted-license records have source links only. Full skill instructions stay at the original source.