Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add little-planet/ascend-tune-lab --skill vllm-ascend-tuninggit clone --depth 1 https://github.com/little-planet/ascend-tune-labWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/little-planet/ascend-tune-lab/vllm-ascend-tuning)<a href="https://agentmods.dev/skills/little-planet/ascend-tune-lab/vllm-ascend-tuning"><img src="https://agentmods.dev/badge/skills/little-planet/ascend-tune-lab/vllm-ascend-tuning/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/little-planet/ascend-tune-lab/vllm-ascend-tuning"><img src="https://agentmods.dev/badge/skills/little-planet/ascend-tune-lab/vllm-ascend-tuning.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00147 | $0.11288 |
| Opus 5 | $0.00073 | $0.05644 |
| Sonnet 5 | $0.00029 | $0.02258 |
| Haiku 4.5 | $0.00015 | $0.01129 |
Grade B, and why
vllm-ascend-tuning scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
sudo apt update Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
wget https://repo.oepkgs.net/ascend/pytorch/vllm/python/py311_bisheng.tar.gz The source is not reproduced here
No licence file
A repository with no LICENSE is all rights reserved by default, so the body is not copied here. The metadata, the measurements and the link are.
What ships with it
37 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/benchmark.md 2.8 KB
- references/cann_hccl.md 1.6 KB
- references/compilation.md 1.4 KB
- references/cpu_binding.md 2.2 KB
- references/env-variables.md 12 KB
- references/features.md 5.3 KB
- references/graph_mode.md 5.7 KB
- references/launch_templates.md 2.0 KB
- references/model_config_guide.md 22 KB
- references/model_configs/DeepSeek-V3.1.yaml 2.0 KB
- references/model_configs/DeepSeek-V3.2.yaml 2.1 KB
- references/model_configs/DeepSeek-V3.yaml 2.0 KB
- references/model_configs/GLM-4.yaml 1.6 KB
- references/model_configs/GLM-5.1.yaml 2.9 KB
- references/model_configs/Kimi-K2.5.yaml 2.1 KB
- references/model_configs/MiniMax-M2.5.yaml 2.1 KB
- references/model_configs/Qwen2.5-VL.yaml 1.7 KB
- references/model_configs/Qwen3-235B-A22B.yaml 1.9 KB
- references/model_configs/Qwen3-30B.yaml 1.8 KB
- references/model_configs/Qwen3-32B.yaml 2.3 KB
- references/model_configs/Qwen3-8B.yaml 1.6 KB
- references/model_configs/Qwen3-Embedding.yaml 1.4 KB
- references/model_configs/Qwen3-MoE-480B.yaml 3.3 KB
- references/model_configs/Qwen3-Reranker.yaml 1.4 KB
- references/model_configs/Qwen3-VL.yaml 1.7 KB
- references/model_configs/Qwen3.5-122B.yaml 2.4 KB
- references/model_configs/Qwen3.5-27B.yaml 2.4 KB
- references/model_configs/Qwen3.5-397B.yaml 2.6 KB
- references/model_matching_guide.md 17 KB
- references/os_level.md 2.3 KB
- references/parallel_strategy.md 3.9 KB
- references/pd_separation.md 6.3 KB
- references/performance_data_summary.md 10 KB
- references/quantization.md 4.5 KB
- references/speculative_decoding.md 4.7 KB
- references/torch_npu.md 1.7 KB
- references/vllm-parameters.md 9.5 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 1,905 lines · 147 tokens per session scan B 713ed051e78a
vllm-ascend-tuning is a skill published in the GitHub repository little-planet/ascend-tune-lab (2 stars, last pushed 22d ago), with no licence file. It adds 147 tokens to every session and 11,288 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it B with 2 findings (asks for root, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
inference-precision-tensor-dump-compare
A debugging tool that captures intermediate tensors, the numerical data passed between model layers, during inference and compares them between GPUs and Ascend NPUs. It supports vLLM and SGLang inference frameworks.
external-gitcode-ascend-ascend-transformer-boost
An index of skills for developing and deploying Huawei's Ascend Transformer Boost library, a set of tools for accelerating transformer-based AI models. It covers the main stages of working with Ascend NPUs and related operator libraries.
ascend-migration-analysis
A feasibility analysis for moving a PyTorch project from CUDA/GPU hardware to Ascend NPU hardware. It scans the code for hardware-specific dependencies and reviews device code, attention mechanisms, custom operations, distributed communication, precision, third-party packages, and compilation.
inferencex-report
Automatically fetch InferenceX benchmark data and generate daily performance reports for LLM inference on various hardware (NVIDIA, AMD, etc.). Supports email delivery, data change detection, and 8k1k sequence length performance analysis. Use when needing to track LLM inference performance trends, compare hardware…
speculative-decoding
Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use when optimizing inference speed (1.5-3.6× speedup), reducing latency for real-time applications, or deploying models with limited compute. Covers draft models, tree-based attention, Jacobi iteration…
distributed-llm-pretraining-torchtitan
Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.