vllm-ascend-correctness-validation git:20260911.9f92063A

Immutable. This exact content is served forever at /api/v1/blob/c0470410f0f93026.

---
name: "vllm-ascend-correctness-validation"
description: "Plan, execute, normalize, and compare vLLM Ascend inference correctness across baseline and candidate code states, eager and graph modes, offline generate or chat, online chat completions, and AISBench task metrics. Use for accuracy validation, token-output comparison, graph-versus-eager checks, deterministic regression testing, or failure classification. Do not use to root-cause an already reproduced graph-only or isolated-operator failure, or for throughput benchmarking, HBM attribution, or profiling-only analysis."
---

<!-- Generated from .agents/skills/vllm-ascend-correctness-validation/SKILL.md. Do not edit. -->

# vllm-ascend-correctness-validation

Read `.agents/skills/vllm-ascend-correctness-validation/SKILL.md` and only the references needed
for the current task. The canonical skill owns the workflow.