air-gapped/skills · .claude/skills/vllm-performance-tuning/SKILL.md

vllm-performance-tuning skillA

vllm-performance-tuning is agent-read markdown (skill) from air-gapped/skills: vLLM performance-tuning operator reference — tuning workflow (baseline → bottleneck → knob → re-bench), fused-MoE kernel autotune (`benchmark_moe.py` generates `E=N,N=M,device_name=X.json` configs), DeepEP all-to-all + expert parallelism + EPLB, CUDA graph modes (FULL_AND_PIECEWISE default), torch.compile AOT + compile cache, scheduler knobs (`--max-num-batched-tokens`, `--max-num-seqs`, `--async-scheduling`), TP/EP/DP/PP decision tree, NCCL/DCGM on H100/H200/B200/GB200, PD disaggregation (Nixl/.

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

How to install

Latest version
mdr add air-gapped/skills/vllm-performance-tuning@git:20260915.4766a48
Exact content
mdr add air-gapped/skills/vllm-performance-tuning@sha256:12decef2e819b36a

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_tpzvanmcusj3lhnr.svg)](https://markdownregistry.com/a/art_tpzvanmcusj3lhnr)

0 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260915.4766a48 latest2026-09-15 4766a48 18,810 BA view · diff
git:20260915.3261b2e2026-09-15 3261b2e 18,751 BA view · diff
git:20260915.8d2641c2026-09-15 8d2641c 18,615 BA view · diff
git:20260915.f93919c2026-09-15 f93919c 18,523 BA view · diff
git:20260811.e17af672026-08-11 e17af67 17,841 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (18810 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

air-gapped/skills · 5 stars · license MIT · pushed 2026-09-16 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_tpzvanmcusj3lhnr
GET https://markdownregistry.com/api/v1/resolve?ref=air-gapped/skills/vllm-performance-tuning
GET https://markdownregistry.com/api/v1/blob/12decef2e819b36a249b4c0053959ab4780f45bd00ec92a4879d0744f60df6d2