model-inference-optimize SKILL.md skillA
model-inference-optimize from majiayu000/spellbook is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add majiayu000/spellbook/model-inference-optimize@git:20261102.d5304c4Pins git:20261102.d5304c4 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- majiayu000/spellbook
- Path
skills/model-inference-optimize/SKILL.md- Versions recorded
- 2
- Latest version
- git:20261102.d5304c4, committed Nov 2, 2026
- First version recorded
- Oct 8, 2026
- Size
- 10,189 bytes
- Repository stars
- 290
- License
- MIT
优化实际模型推理链路，将正确性对齐、分段 profiling、显存与数据搬运、TensorRT/ONNX/PyTorch 后端、attention/kernel、FP8/compile、缓存与少步采样、质量回归、GPU 成本和服务验收串成同一实验闭环。当用户要求推理提速、降低显存或 GPU 成本、复现模型效果、定位 GPU 利用率低、优化图像/视频/扩散模型或自托管 LLM 时使用，提供 model inference performance optimization 的脱敏案例。纯 API 调用、提示词或 token 用量优化，以及电脑卡顿，不进入本技能的模型推理实验流程。.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# 模型推理优化系统 把明确的推理工作负载优化到可验证的结果。沿用同一任务定义、实验记录和验收标准，按瓶颈选择专项，不把所有优化技术轮流试一遍。 ## 入口与资料 | 当前问题 | 从哪里开始 | 按需读取 | |---|---|---| | 模型尚未跑通、迁移后效果不对 | 任务 → 正确性对齐 | [专项方法](references/playbook.md) 的「复现与后端迁移」 | | 延迟高、GPU 利用率低、显存不足 | 任务 → 基线 → profile | 专项方法的「测量」「显存与数据流」 | | 换 attention/kernel、FP8 或 compile | 先证明热点和实际运行路径 | 专项方法的「算子与编译」 | | 减少 steps、换蒸馏模型或近似 decoder | 先定义允许的质量变化 | 专项方法的「近似推理」 | | 复用同类优化经验 | 按当前瓶颈选案例 | [脱敏案例](references/cases.md)，只读相关小节 | | 已有优化要接入服务或证明真实收益 | 质量回归 → 完整请求 → 服务验收 | 专项方法的「质量评测」「服务与成本」 | 需要记录时复制 [实验模板](references/experiment-template.md) 到本轮产物目录，删去不适用的字段；普通 bug 修复可缩成一张对比表。 模型质量/性能对比沿用本技能。纯托管 API、Prompt、token/cache 命中或 agent 工具调用优化不进入下述 GPU 实验流程：先检查当前可用技能列表，仅在外部技能 `llm-app-optimize` 已安装时转交；未安装时，在当前对话中固定代表请求，记录端到端延迟、用量/费用和输出质量，一次修改一个因素，并复用应用已有的回归检查验收。 全产品竞品研究仅在外部技能 [`peerscope`](https://github.com/majiayu000/peerscope) 已安装时转交；未安装时，在当前对话中明确产品对比问题，从官方资料收集功能、价格和使用限制，附来源与日期，并将未实测能力标为待验证。这两个外部技能均不是 Spellbook 安装前提，不要求为完成当前任务安装它们。 ## Operating Contract - **Direct actions:** 在请求范围内读代码、分析现有日志、修改本地实现并运行已授权的检查和实验。 - **Escalate before:** 超出既有授权的收费运行、凭证修改、占用其他任务设备、生产变更或发布。已有授权覆盖时继续执行。 - **Evidence-backed pushback:** profile 显示拟优化部分占比很低或任务规格被改变时，给出 trace/输出证据和更小的替代实验。 …
Read the whole file at its exact version.
How to install
mdr add majiayu000/spellbook/model-inference-optimize@sha256:8b21e3b7756cd41bThe line at the top of this page pins the label git:20261102.d5304c4, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_rj4licnhdvu5f7ft)
4 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 2 versions of this file, between Oct 8, 2026 and Nov 2, 2026, about one every 26 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261102.d5304c4 latest | 2026-11-02 | d5304c4 | 10,189 B | A | view · diff |
| git:20261008.6310f81 | 2026-10-08 | 6310f81 | 10,146 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (10189 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
majiayu000/spellbook · 290 stars · license MIT · pushed 2026-11-04 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_rj4licnhdvu5f7ft GET https://markdownregistry.com/api/v1/resolve?ref=majiayu000/spellbook/model-inference-optimize GET https://markdownregistry.com/api/v1/blob/8b21e3b7756cd41be39d8a1b97efb8b2fa2e1cee9300a2091a967f2a4370b34a
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.