7×24 chaos testing for RobustMQ. Injects broker-kill and network-delay faults, validates SDK client resilience across Python/Go/Rust/Java, and publishes a Markdown + JSON report to GitHub after each run.

Turn a raster logo (PNG/JPG/WebP/screenshot) into a clean minimal SVG with edge smoothness as the primary hard gate and IoU optimized as high as reasonably possible without a fixed global threshold, then into a choreographed logo animation delivered as standalone JS-rendered HTML, applying Disney's 12 animation principles. Use when asked to animate a logo, build a logo reveal / splash screen / brand intro, convert a logo image into animated SVG or HTML, add motion to a vectorized mark, or create loading/idle/hover motion for a brand mark. v2: also handles self-crossing draw-on choreography (split-fill, exact easing subdivision, tip glint), closed variable-width ribbon fitting, and quantitative motion QA (easing probe, ink-delta continuity sweep).

对目标话题进行初步调研，生成调研outline。用于学术调研、benchmark调研、技术选型等场景。

将deep调研结果汇总为markdown报告，覆盖所有字段，跳过不确定值。

读取调研outline，为每个item启动独立agent进行深度调研。禁用task output。

向现有调研outline补充items（调研对象）。

向现有调研outline补充字段定义。

Conduct preliminary research on a topic and generate research outline. For academic research, benchmark research, technology selection, etc.

Summarize deep research results into markdown report, cover all fields, skip uncertain values.

Read research outline, launch independent agent for each item for deep research. Disable task output.

Add items (research objects) to existing research outline.

Add field definitions to existing research outline.

Conduct preliminary research on a topic and generate research outline. For academic research, benchmark research, technology selection, etc.

Summarize deep research results into markdown report, cover all fields, skip uncertain values.

Read research outline, launch independent agent for each item for deep research. Disable task output.

Add items (research objects) to existing research outline.

Add field definitions to existing research outline.

Run the Nemotron-3 Ultra Text2SQL LoRA fine-tuning tutorial (NeMo Megatron-Bridge) end-to-end for the user on their SLURM cluster: data prep, distributed checkpoint conversion, and packed LoRA fine-tuning of the 550B hybrid Mamba-Transformer MoE, ending at a saved adapter. Use when the user wants to run this cookbook, fine-tune Nemotron-3 Ultra with LoRA, or adapt the notebook to their own cluster.

Reference desk for NVIDIA Nemotron 3 Ultra (550B-A55B) — architecture, NVFP4 pretraining, SFT, MOPD (multi-teacher on-policy distillation), MTP boosting, quantization, inference. Use when the user asks facts about Ultra rather than building a pipeline.

Reference desk for NVIDIA Nemotron 3 Super — architecture, training data, recipes (pretrain/SFT/RL/eval/quantization), and deployment notes. Use when the user asks facts about Super3 rather than building a pipeline.