Results for “latency-optimization”

Reduce per-token latency on Jetson vLLM servers by appending speculative decoding configuration, with guidance on when to enable and how to benchmark the improvement.

`analysis-agent`/`task-agent`/`review-agent`: use when latency, throughput, bundle, memory, CPU, query, rendering, or resource cost needs a budget; skip without performance risk.

Measures inference latency of binarized, 8-bit, and 32-bit convolutional layers on edge devices to evaluate the efficiency and speedup of the Larq Compute Engine framework compared to standard implementations.

Benchmark Jetson LLM/VLM serving performance across vLLM, llama.cpp, and Ollama with structured JSON output.

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency on NVIDIA GPUs (A100/H100).

Shares memory between agents at the representation level by compacting the orchestrator's KV cache for efficient worker handoff, reducing token costs without summarization or retrieval.

The layers trade timeliness against definitiveness.

Optimize context usage for the session by reducing bloat, improving token efficiency, and focusing on relevant areas. Use when context is growing large, responses are slow, or switching between unrelated tasks. Covers context analysis, compacting, targeted pruning, MCP-first strategies, and focused scoping by area.

Quantize LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss using bitsandbytes. Supports INT8, NF4, FP4 formats, QLoRA training, and 8-bit optimizers.

Reduce OpenClaw token usage and API costs through smart model routing, heartbeat optimization, budget tracking, and native 2026.2.15 features (session pruning, bootstrap size limits, cache TTL alignment). Use when token costs are high, API rate limits are being hit, or hosting multiple agents at scale. The 4 executable scripts (context_optimizer, model_router, heartbeat_optimizer, token_tracker) are local-only — no network requests, no subprocess calls, no system modifications. Reference files (PROVIDERS.md, config-patches.json) document optional multi-provider strategies that require external API keys and network access if you choose to use them. See SECURITY.md for full breakdown.

Designing and evaluating AI/ML systems: prompt engineering, RAG design, LLM application patterns, AI safety, evaluation frameworks, MLOps, cost optimization. Use for AI pipelines or eval harnesses.

Improves AI skills for activation, clarity, and cross-model reliability through benchmarking, salience tuning, and regression triage.

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

Provides guidance and code for production AI workflows including RAG pipelines, vector databases, embedding indexing, prompt engineering, cost estimation, semantic caching, and quality evaluation.

Evaluates computational throughput and real-time efficiency of embedded CPU and GPU platforms by measuring peak FLOPS via a matrix rotation kernel and assessing inference latency and power consumption on a robotic vision pipeline.

Transform the agent into a performance engineer. Apply methodologies for measuring, profiling, and optimizing code (caching, algorithm complexity, resource usage).

Complexity-based task routing with Q-Learning optimization, Agent Booster WASM fast-path, and Mixture-of-Experts model selection.

Expert prompt optimization system for the prompts INSIDE an AI product you are building — system prompts, LLM feature prompts, chatbot/agent instructions. Use when the user wants to write or improve a system prompt for an AI feature they're shipping, review/critique an LLM prompt, apply prompt-engineering techniques (chain-of-thought, few-shot, structured output, hard constraints) to a product prompt, or optimize cost/latency of a production prompt. Do NOT use this to clarify or structure the user's own vague request to Claude Code — that is `prompt-clarifier`'s job, not this skill's.