Stats

Actions

Available In

Help us improve

Share bugs, ideas, or general feedback.

Skill

Use it when

llm-serving-auto-benchmark

You need a fair, bounded serving benchmark search for SGLang, vLLM, TensorRT-LLM, or another OpenAI-compatible stack.

llm-serving-capacity-planner

You need to explain SGLang or vLLM startup memory, KV cache budget, request capacity, or OOM pressure from logs.

llm-torch-profiler-analysis

You need a three-table profiler report that keeps extend/prefill and decode evidence separate.

llm-pipeline-analysis

You need forward-pass, layer, and kernel-level timing from a torch profiler trace, including anchor boundaries and Perfetto ranges.

model-compute-simulation

You need operator shapes, FLOPs, MFU estimates, kernel-to-op mapping, or parallelism what-if analysis for an LLM serving shape.

sglang-humanize-review

You need SGLang code-review findings grounded in 2024-2025 human review threads, including inline code context, comments, and discussions.

sglang-sota-humanize-loop

You want one model-level Humanize RLCR loop that owns gap decisions, profiler triage, required layer-pipeline deep dives, SGLang patches, optional ncu-report-skill evidence, and real-model revalidation after the fixed fair benchmark.

vllm-sota-humanize-loop

You want one model-level Humanize RLCR loop that owns gap decisions, profiler triage, required layer-pipeline deep dives, vLLM patches, optional ncu-report-skill evidence, and real-model revalidation after the fixed fair benchmark.

sglang-prod-incident-triage

You need to turn queue growth, timeouts, wrong outputs, crashes, or distributed stalls into a replay and next debug step.

model-architecture-diagram

You need original public architecture diagrams for popular LLM, VLM, MoE, OCR, and diffusion model families.

AI-Infra-Auto-Driven-SKILLS

Agent-ready playbooks for LLM serving benchmarks, capacity planning, torch-profiler triage, pipeline analysis, compute simulation, SGLang/vLLM optimization, human code review, production incidents, and model PR intelligence.

This repository is built for AI infrastructure engineers who want agents to do real work, not recite generic prompts.

It gives an agent the operational memory needed to benchmark SGLang, vLLM, and TensorRT-LLM fairly; explain serving capacity from startup logs; split prefill and decode profiler evidence; inspect traces at layer and kernel level; estimate operator FLOPs and MFU; review SGLang patches against real maintainer discussion patterns; run Humanize-governed SGLang and vLLM SOTA loops; triage SGLang production incidents from a replay; and keep model-family optimization history close to the code that actually changed.

For standalone kernel campaigns and kernel evidence tools, see the sibling project KernelPilot.

If this saves you one stale model-support assumption, one misleading profiler trace, or one late-night benchmark loop, a star helps more AI-infra engineers find it.

Core Skills

Skill	Use it when
`llm-serving-auto-benchmark`	You need a fair, bounded serving benchmark search for SGLang, vLLM, TensorRT-LLM, or another OpenAI-compatible stack.
`llm-serving-capacity-planner`	You need to explain SGLang or vLLM startup memory, KV cache budget, request capacity, or OOM pressure from logs.
`llm-torch-profiler-analysis`	You need a three-table profiler report that keeps `extend/prefill` and `decode` evidence separate.
`llm-pipeline-analysis`	You need forward-pass, layer, and kernel-level timing from a torch profiler trace, including anchor boundaries and Perfetto ranges.
`model-compute-simulation`	You need operator shapes, FLOPs, MFU estimates, kernel-to-op mapping, or parallelism what-if analysis for an LLM serving shape.
`sglang-humanize-review`	You need SGLang code-review findings grounded in 2024-2025 human review threads, including inline code context, comments, and discussions.
`sglang-sota-humanize-loop`	You want one model-level Humanize RLCR loop that owns gap decisions, profiler triage, required layer-pipeline deep dives, SGLang patches, optional `ncu-report-skill` evidence, and real-model revalidation after the fixed fair benchmark.
`vllm-sota-humanize-loop`	You want one model-level Humanize RLCR loop that owns gap decisions, profiler triage, required layer-pipeline deep dives, vLLM patches, optional `ncu-report-skill` evidence, and real-model revalidation after the fixed fair benchmark.
`sglang-prod-incident-triage`	You need to turn queue growth, timeouts, wrong outputs, crashes, or distributed stalls into a replay and next debug step.
`model-architecture-diagram`	You need original public architecture diagrams for popular LLM, VLM, MoE, OCR, and diffusion model families.

SGLang SOTA Performance Loop

Model PR History Catalog

The model optimization layer is now one knowledge base: model-pr-optimization-history. It contains 58 PR-driven history dossiers and a small query helper. These are not per-model runbook skills; they preserve diff-backed model evolution records for SGLang and vLLM so SOTA loops can read prior source and PR evidence before patching.

Help us improve

Find plugins for your project

Help us improve

ai-infra-auto-driven-skills

Popularity

What's Inside

Help us improve

Health & Quality

Confidence

README

AI-Infra-Auto-Driven-SKILLS

Core Skills

SGLang SOTA Performance Loop

Model PR History Catalog

Similar Plugins

vllm-skills

nemo-evaluator-skills

langfuse-pack

skill-optimizer

superml

tensorzero

More by BBuf

humanize

AI-Infra-Auto-Driven-SKILLS

Core Skills

SGLang SOTA Performance Loop

Model PR History Catalog