Skills

Skills are ready-to-run expert methods for Zeplik, the AI assistant that works across every model in one chat. Browse 445 skills below. Ask in plain language and Zeplik applies the right method for you, or open any skill to try it directly.

Featured skills

• Agent MemoryMemory and context management for LLM agents — short/long-term memory stores, conversation persistence, context-window strategies (summarization, trimming, retrieval) and prompt caching. Use for "give my agent memory / manage context"; for vector stores see vector-databases.
• AI Safety GuardrailsSafety and moderation for LLM apps — Constitutional AI, Llama Guard input/output moderation, and NeMo Guardrails runtime rails. Use for "add safety/moderation/guardrails to my LLM app"; for evaluating safety see llm-evaluation-harnesses.
• Autonomous Agent PatternsFramework-agnostic design patterns for autonomous coding agents — goal decomposition, planning, parallel/multi-agent orchestration and operational modes. Use for "how should I architect an autonomous agent"; for concrete frameworks see ai-agent-frameworks.
• Gemini CLIRun Gemini CLI (Gemini 3 Pro) for code/plan review and huge >200k-token context analysis in the terminal
• InterpretabilityMechanistic interpretability of neural networks — TransformerLens, NNsight remote access, sparse autoencoders (SAELens), and causal interventions (pyvene). Use for "probe/intervene on model internals" or "train an SAE"; for architecture basics see llm-architectures.
• LLM App PatternsProduction patterns for LLM applications — RAG architecture, embeddings, LLMOps, and end-to-end app design. Use for "architect a production LLM/RAG app"; for the vector store see vector-databases, for agents see ai-agent-frameworks.
• LLM ArchitecturesTransformer and alternative model architectures — attention/transformers, long-context (RoPE, YaRN, ALiBi), Mamba/state-space, RWKV, and nanoGPT-scale implementations. Use for "understand/implement a model architecture"; for training it see llm-training-frameworks.
• LLM Eval HarnessesBenchmarking and evaluating LLMs and agents — lm-evaluation-harness, code-model benchmarks (HumanEval/MBPP), NeMo Evaluator, agent behavioral testing and test-plan authoring. Use for "benchmark/evaluate a model or agent"; for tracing runs see llm-observability.
• LLM ObservabilityTracing, monitoring and experiment tracking for LLM and ML apps — Langfuse, LangSmith, Phoenix, MLflow, Weights & Biases, TensorBoard. Use for "trace/monitor/log my LLM app or training runs"; for evaluation benchmarks see llm-evaluation-harnesses.
• LLM QuantizationCompressing and accelerating LLMs — AWQ, GPTQ, GGUF/llama.cpp, bitsandbytes, HQQ, plus distillation, pruning, model merging, FlashAttention and speculative decoding. Use for "quantize/compress/speed up a model"; for high-throughput serving see llm-serving.
• LLM ServingHigh-throughput LLM inference serving — vLLM PagedAttention, SGLang, TensorRT-LLM, and llama.cpp for CPU/Apple Silicon. Use for "serve/deploy a model for inference at scale"; for shrinking the model first see llm-quantization.
• LLM Training FrameworksTraining and fine-tuning large models — distributed pretraining (Megatron, FSDP, torchtitan), RLHF and MoE training, HF Transformers/Axolotl/NeMo, tokenizers and data pipelines. Use for "train/fine-tune with X"; for serving see llm-serving, quantizing see llm-quantization.