sr-search-record SKILL.md skillA
sr-search-record from qiaomiaojoe/ai-academic-workflow is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add qiaomiaojoe/ai-academic-workflow/sr-search-record@git:20261023.c385d52Pins git:20261023.c385d52 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- qiaomiaojoe/ai-academic-workflow
- Path
skills/sr-search-record/SKILL.md- Versions recorded
- 5
- Latest version
- git:20261023.c385d52, committed Oct 23, 2026
- First version recorded
- Aug 4, 2026
- Size
- 26,361 bytes
- Repository stars
- 10
- License
- a license file GitHub does not identify
综述文章的检索与筛选记录（乔淼PhD · AI学术训练营 · AI学术工作流 · 综述文章工作台 阶段②）。**Snyder 三型通用**——系统 / 半系统·叙事 / 整合式共用一套内核。AI 全自动跑完：概念块拆解 → 检索式多轮自动校准（种子召回 + 精度守卫）→ 全量抓取去重 → **按相关性排序截断出候选集** → 摘要筛选（**摘要层定稿，不做全文筛选**）→ include 收敛到目标带 → 引文追踪补漏 → Zotero 入库 → 产出论文可直接贴的三张表（检索记录表 S1–Sn / 纳入排除标准表 / 选择流程数字）+ 待人工复核清单。**全程只有一个硬停点**，其余一律降级继续。检索引擎用 OpenAlex 开放 API（免登录；翻页要免费 API key，随 skill 附 scripts/oa_search.py），医学题目补 Europe PMC（MeSH）。AI 只出筛选建议不当第二评审。Trigger on："跑综述的文献检索"、"自动检索并筛选文献"、"建综述语料库"、"生成检索记录表和 PRISMA 数字"、…
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
# SR Search Record · 综述文章的检索与筛选记录 把综述文章的检索环节做成**可复现、数字能对账、且一口气跑得完**的流水线。**Snyder 三型共用这一套**（历史原因名字里带 sr，不只服务 systematic review）。AI 干 80%（设计检索式、多轮校准、全量抓取、排序截断、摘要筛选建议、记账出表），剩下 20%（最终纳入裁决、边界记录、学科术语补漏）交给人。 **产出不是"搜到的一堆文献"，是论文方法节能直接贴的三张表 + 一个 60–100 篇的纳入集。** ## 摘要层定稿模式（本 skill 的核心设定） **判定只有一层：题名 / 摘要。`ai_decision` 就是纳入决定，没有全文筛选层。** - 筛选阶段**不获取全文**。全文在 ③ 成文阶段（`review-synthesis` / `review-draft`）best-effort 获取，**拿不到不改变纳入集**，只影响 `evidence_level`。 - `ft_decision` / `ft_reason` 两列已退休。台账里不再有它们。 - 这是**明示的方法学让步，不是无损优化**，三型都适用： - 摘要经常不写清关键闸门（样本、设计、是否报告了某类结果），只靠摘要判会同时放进假阳性、漏掉假阴性。 - **PRISMA 要求全文评估**。systematic 型用本模式，产出里必须写明这条偏离，投稿时大概率被要求补做。 - RAMESES（半系统）与 Torraco（整合式）不强制全文评估层，与本模式不冲突。 - **产出里一律如实写**："筛选在题名/摘要层完成，未进行全文合格性评估"。**不得把流程图标成完整 PRISMA**，systematic 型的流程图要标注该偏离。 ## 规模链（三个数必须一致，流程才闭得上） | 环节 | 目标 | 谁保证 | |---|---|---| | 检索式校准后命中 | **300–2,000** | 第 2 步，种子召回 + 精度守卫 | | 排序截断后候选集 | **300 条**（可扩到 600 / 900） | 第 4 步 `rank --shortlist` | | 摘要筛选后 include | **60–100** | 第 4 步收敛闭环 | 实测参考（在一个 7,660 条的真实语料库上验证）：截断线 300 时候选集里 include 密度约 23–25%，**预计产出 70–80 篇**，落在目标带内。 …
Read the whole file at its exact version.
How to install
mdr add qiaomiaojoe/ai-academic-workflow/sr-search-record@sha256:a43a6d86b1af337bThe line at the top of this page pins the label git:20261023.c385d52, 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_xzetm2sqpvrgaans)
0 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 5 versions of this file, between Aug 4, 2026 and Oct 23, 2026, about one every 20 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261023.c385d52 latest | 2026-10-23 | c385d52 | 26,361 B | A | view · diff |
| git:20261015.26ae0d2 | 2026-10-15 | 26ae0d2 | 26,160 B | A | view · diff |
| git:20261015.61b50d3 | 2026-10-15 | 61b50d3 | 25,005 B | A | view · diff |
| git:20260806.db9e8fc | 2026-08-06 | db9e8fc | 24,974 B | A | view · diff |
| git:20260804.ad7c276 | 2026-08-04 | ad7c276 | 25,075 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 (26361 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
qiaomiaojoe/ai-academic-workflow · 10 stars · license a license file GitHub does not identify · pushed 2026-11-03 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_xzetm2sqpvrgaans GET https://markdownregistry.com/api/v1/resolve?ref=qiaomiaojoe/ai-academic-workflow/sr-search-record GET https://markdownregistry.com/api/v1/blob/a43a6d86b1af337bf1923dc0c19400eb041d4934b50ff8ca3c711fd5b304830b
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.