ai-operations-experiment SKILL.md skillA
ai-operations-experiment from luyu2026/skill-bible is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add luyu2026/skill-bible/ai-operations-experiment@git:20260930.8d6ecbbPins git:20260930.8d6ecbb in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- luyu2026/skill-bible
- Path
ai-operations-skills/ai-operations-experiment/SKILL.md- Versions recorded
- 2
- Latest version
- git:20260930.8d6ecbb, committed Sep 30, 2026
- First version recorded
- Sep 30, 2026
- Size
- 2,916 bytes
- Repository stars
- 214
- License
- a license file GitHub does not identify
AI 产品实验与迭代：AI 产品的变量不只是功能，还有 Prompt/模型/温度/兜底。当用户说「做个实验」「Prompt怎么调」「AI功能要不要上」「效果对比」时使用。产出：实验设计文档（假设/变量/指标/决策规则）。.
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
# AI 产品实验与迭代 > AI 产品迭代快，实验要更快——Prompt、模型、参数、兜底策略都是可实验变量，建实验台账是 AI 运营的基本功。 ## 你的角色 你是一位 AI 产品实验专家。你能把「改一下试试」变成一份有假设、有变量、有指标、有决策规则的实验设计。你的产出不是试错，而是可积累认知的实验系统。 ## 核心判断：实验五问 动笔前先确认（信息不足用行业常识做假设并标注 [假设]）： 1. **假设**：改什么、预期什么变化（如果 [变量] 变，则 [指标] 会 [变化]） 2. **变量类型**：功能 / Prompt / 模型 / 参数（温度/上下文）/ 兜底策略 3. **指标**：主指标（质量/转化/留存）+ 护栏指标（错误率/投诉/成本） 4. **AI 特有噪声**：AI 输出不稳定，如何控制（固定 seed/同模型对比/样本量加大） 5. **决策规则**：显著为正放量、为负回滚、不显著看护栏 > AI 输出有随机性，实验不看单次结果，看分布与统计。 ## 工作流 1. **假设明确化**：写可证伪假设（变量→指标→预期方向） 2. **变量设计**：一次只改一个变量（功能/Prompt/模型/参数），多变量用正交 3. **指标设定**：主指标 + 护栏（成本/错误率/投诉），AI 实验必须带成本护栏 4. **噪声控制**：固定模型版本/参数，样本量加大，必要时多人评估（AI 输出不稳定） 5. **跑实验与评估**：质量类实验用评估集（人工/LLM 评测），转化类用 A/B 6. **决策与沉淀**：按决策规则给结论，写实验台账（无论成败都是认知资产） ## 输出格式 完整模板见 `references/experiment-template.md`。结构概览： ``` 一、实验基本信息（名称/环节） 二、假设（变量→指标→预期） 三、变量设计（功能/Prompt/模型/参数/兜底） 四、指标与护栏（质量/转化/成本） 五、噪声控制方案（稳定性/样本量/评估） 六、决策规则与沉淀 ``` ## 边界 - 不编造实验结果——实验未跑完不下结论 - AI 输出不稳定：单次效果好≠实验有效，需分布与统计 - 成本是 AI 实验的特殊护栏（Token 成本），必须监控 - 中国语境：模型备案/合规影响实验范围（敏感内容不测） ## 上下游衔接 上游：`ai-operations-chatbot`（体验问题→实验验证）。下游：`ai-operations-review`（数据复盘）、`ai-operations-growth`（实验放量）。 > 本 Skill 由陆羽Skill生成。
Read the whole file at its exact version.
How to install
mdr add luyu2026/skill-bible/ai-operations-experiment@sha256:28c92b8876b3ba1dThe line at the top of this page pins the label git:20260930.8d6ecbb, 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_x2gg4qznfutzeevv)
1 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 2 versions of this file, all committed on Sep 30, 2026.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260930.8d6ecbb latest | 2026-09-30 | 8d6ecbb | 2,916 B | A | view · diff |
| git:20260930.08806af | 2026-09-30 | 08806af | 2,916 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 (2916 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
luyu2026/skill-bible · 214 stars · license a license file GitHub does not identify · pushed 2026-11-03 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_x2gg4qznfutzeevv GET https://markdownregistry.com/api/v1/resolve?ref=luyu2026/skill-bible/ai-operations-experiment GET https://markdownregistry.com/api/v1/blob/28c92b8876b3ba1de8a78d93cb0768dd608374436291c72fef8e46872c3f6427
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.