434 skills · sorted by relevance
Build and troubleshoot applications that move DEFLATE compression and decompression, or LZ4 decompression, to DOCA-supported hardware. Guidance covers device capabilities, buffer sizes, memory permissions, validation, and deciding when CPU processing is more suitable.
Configure prompt compression, which shortens text sent to AI models, through the OmniRoute command line. Adjust RTK filters, Caveman rules, and combined compression modes, then preview the results using real prompts.
Configure compression for command output, prose, and chat history before sending them to AI providers. Choose suitable engines, adjust language rules, and preview compressed text and reported savings before enabling a mode.
Reduce image file sizes by compressing individual files or whole folders into WebP, PNG, or JPEG. The skill selects an available compression tool automatically and lets you control quality, include subfolders, and choose whether to keep originals.
Compress memory files and task lists into shorter wording to reduce input tokens. Keeps code and document structure intact, saves a readable backup separately, and sends the text to the configured model provider.
Rewrite lengthy prompts, tool descriptions, and agent instructions into compact text while preserving requirements, exceptions, and exact values. The process records removed content and compares token counts, the units models use to measure text length.
Reduce LLM weight usage through pruning, which removes selected model parameters. Explore Wanda and SparseGPT for compression without retraining, choose hardware-friendly patterns for faster inference, and compare the compressed model's accuracy with the original.
Manage how context, the text supplied to an AI, is compressed with RTK filters. Preview compression on real prompts, validate and install filter files, and inspect filter loading diagnostics.
Review website images and reduce file sizes while limiting visible quality changes. Check compression, formats, display dimensions, loading behavior, and delivery settings, then recommend suitable tools or build automation and verification in browser DevTools.
Compress large language models to 4-bit weights to reduce GPU memory use and accelerate inference. Covers loading compressed models, configuring quantization, and using Transformers or vLLM, with guidance on hardware requirements and backend choices.
Improve Astro site loading with critical CSS, the styles needed for the first screen, plus file compression, font loading adjustments, main image preloading, and deferred third-party scripts. Includes integration settings and a layout example.
Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman-compress FILEPATH or "compress memory file"
The English summary is being prepared. Showing the original description.
Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman-compress FILEPATH or "compress memory file"
The English summary is being prepared. Showing the original description.
Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman:compress <filepath> or "compress memory file"
The English summary is being prepared. Showing the original description.
Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman:compress <filepath> or "compress memory file"
The English summary is being prepared. Showing the original description.
Compress an oversized SKILL.md to under 200 lines without losing effectiveness. Load when a skill exceeds 200 lines, when AGENTS.md triggers compression after a skill edit, or when the user asks to compress, shrink, slim down, or optimize a skill. Also triggers on "this skill is too long", "reduce skill size", "make this skill shorter". Applies to all skills including meta skills — the 200-line rule has no exceptions. Preserves hard gates, gotchas, output format, routing triggers, and at least one example. When genuinely CORE content cannot be compressed away, invokes split-skill instead of degrading the skill.
The English summary is being prepared. Showing the original description.
Check whether HTML, JavaScript, and CSS are delivered with Gzip or Brotli compression. Use measured loading problems and response headers to guide server or CDN changes that reduce the amount of data downloaded.
Process video and audio with FFmpeg for editing and publishing. Convert formats, resize and compress files, extract audio, trim clips, and change playback speed. Includes recipes for preparing Remotion assets and exporting videos for social platforms and websites.
Edit local video files with ffmpeg, a command-line media tool. Trim and join clips, resize for social platforms, adjust playback speed, extract or replace audio, add image overlays, and export compressed or converted files.
Design structured summaries for long agent sessions that preserve goals, file changes, decisions, and next actions. Choose compression methods and timing, then test whether the agent can continue accurately without repeatedly retrieving information it already processed.
Guides LLM compression through pruning, which removes selected model weights. Covers Wanda and SparseGPT methods without retraining, accuracy comparisons, and structured weight patterns that can accelerate inference on compatible hardware.
Guide knowledge distillation, where a smaller language model learns from a larger teacher. Covers training from teacher responses or prediction scores, adjusting training settings, and comparing outputs to support model compression and specialized training.
Compress large language model weights to four bits with AWQ, a method that uses model activation patterns to limit quality loss. Follow workflows for quantization, loading compressed models, and serving them with vLLM using less GPU memory.
Measure throughput, latency, or maximum bandwidth for DOCA libraries using doca_bench on a host or BlueField Arm. Capture commands and environment details for repeatable comparisons, check run stability, and investigate configuration or measurement problems.