AGENTS.md@src/eval-workflowsA
AGENTS.md@src/eval-workflows from lizhiyao/oh-my-knowledge is an AGENTS.md file: the standing instructions coding agents read in that repo.
mdr add lizhiyao/oh-my-knowledge/src/eval-workflows/AGENTS.md@git:20260906.26cd95aPins git:20260906.26cd95a in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 14 of 14 checks passed
- Kind
- AGENTS.md file
- Repository
- lizhiyao/oh-my-knowledge
- Path
src/eval-workflows/AGENTS.md- Versions recorded
- 4
- Latest version
- git:20260906.26cd95a, committed Sep 6, 2026
- First version recorded
- Sep 3, 2026
- Size
- 1,829 bytes
- Repository stars
- 24
- License
- MIT
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
# Evaluation Workflows 领域规则 本文件补充仓库根 `AGENTS.md`，适用于 `src/eval-workflows/`。本层负责产品输入、评分与统计声明、结果投影、持久化及评测工作流；执行能力通过注入的 Evaluation Runtime 接口接入，具体宿主装配位于 `hosts/`，由 CLI／DSH 共用。`hosts` 是独立装配边界；其它子域不得反向导入它，不能因同属本目录而自由互相依赖。 ## 领域不变量 - OMK 产品工作流的评分类 prompt 位于 `instruments/prompts/`；公共 Rubric Judge prompt 的单一来源位于 `eval-runtime/judges/rubric-prompt.ts`，本层仅复用。所有 prompt 均由 `test/measurement-governance/prompt-registry.ts` 与 `prompt-registry-freeze.test.ts` 登记、冻结。 - 五层评分管道的语义固定为 assertion／llm／judge／dimension／composite。Bootstrap CI 和 Krippendorff alpha 公式不得静默变化。 - `--no-debias-length` 只关闭 rubric 评委的长度去偏 prompt 变体；排版和语气中性化始终开启。 - 工作流可以编排副作用，但必须通过显式 adapter 接入，不能把宿主状态泄漏进 Core 契约。 ## Code Review Rules - 必须拦截评分类 prompt 的非预期字节漂移，以及只为让 CI 通过而更新冻结 hash。安全路径：保持冻结文本；确需改变评分构念时，同步版本、回归证据和 `BREAKING-COMPARABILITY` 说明。 - 必须拦截五层评分、统计公式、缺失证据处理或 verdict 聚合语义的静默变化。安全路径：显式版本化并用固定 fixture 证明新旧结果边界。 - 必须拦截「固定模型、只变知识载体」实验隔离被破坏，或失败证据被当作成功分数。安全路径：在输入编译和 adapter 边界校验实验变量，缺失关键证据时 fail closed。
Read the whole file at its exact version.
How to install
mdr add lizhiyao/oh-my-knowledge/src/eval-workflows/AGENTS.md@sha256:b329dda9ac82efc3The line at the top of this page pins the label git:20260906.26cd95a, 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_udzsvzmwtk7eiufs)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 4 versions of this file, between Sep 3, 2026 and Sep 6, 2026, about one every day.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260906.26cd95a latest | 2026-09-06 | 26cd95a | 1,829 B | A | view · diff |
| git:20260905.c155b6e | 2026-09-05 | c155b6e | 1,699 B | A | view · diff |
| git:20260903.03bb38c | 2026-09-03 | 03bb38c | 1,593 B | A | view · diff |
| git:20260903.08c59f3 | 2026-09-03 | 08c59f3 | 1,461 B | A | view |
Audit of the latest version
- pass: Size between 200 bytes and 200 KB (1829 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
lizhiyao/oh-my-knowledge · 24 stars · license MIT · pushed 2026-11-08 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_udzsvzmwtk7eiufs GET https://markdownregistry.com/api/v1/resolve?ref=lizhiyao/oh-my-knowledge/src/eval-workflows/AGENTS.md GET https://markdownregistry.com/api/v1/blob/b329dda9ac82efc396298865b408f5dda905609378bcd90259d0241346faf421
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: how to write a good AGENTS.md · AGENTS.md vs CLAUDE.md · which agent file does what.