Jiujiang shuangzheng product material organizer
为九江双蒸把客户提供的产品素材文件夹或 zip 压缩包整理入 Obsidian 产品素材库。用于“整理产品素材包”“导入九江双蒸 01.4_产品素材库”“按图片、视频、文档分类产品素材”“把图片分到产品实拍、好评截图、用户晒单、用户咨询”等任务;支持中文 zip 路径修复、素材盘点、按文件类型和产品生成预案、确认后复制入库和维护统一素材入库总表。不用于生成卖点卡、产品资料卡、朋友圈文案、合规审查、删除素材、覆盖原文件或低置信强制分类。From its SKILL.md
npx -y skills add NewbieCheng/company-skills-marketplace --skill jiujiang-shuangzheng-product-material-organizerAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 20 days oldThe repository was created 20 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
6.2 KB, ~2.0k tokens by cl100k_base, as published. Nobody here has run it
九江双蒸产品素材入库分类
核心原则
把素材整理到:
九江双蒸知识库v1.0/一、私域库/01_产品库/01.4_产品素材库/
目录主结构固定为:
01.4_产品素材库/
├── 图片类素材/
│ └── <产品名>/
│ ├── 00_待人工确认/
│ ├── 01_产品实拍/
│ ├── 02_好评截图/
│ ├── 03_用户晒单/
│ └── 04_用户咨询/
├── 视频类素材/
│ └── <产品名>/
├── 文档类素材/
│ └── <产品名>/
├── 00_待人工确认/
└── 99_入库索引/
先判断文件类型,再判断产品。只有图片类素材继续细分 产品实拍 / 好评截图 / 用户晒单 / 用户咨询;视频类和文档类不做这些用户场景细分。
99_入库索引/素材入库总表.md 是唯一共享总表。图片、视频、文档每次入库都追加到同一份总表;不要按文件类型另建索引表。
必读资源
- 分类前读取 references/category-rules.md。
- 生成目录或路径前读取 references/target-structure.md。
- 识别产品名或处理别名时读取 references/product-name-map.md。
工作流
1. 盘点输入
输入可以是文件夹或 zip。先运行只读盘点:
python3 scripts/inspect_materials.py <输入路径> --output-dir <本次运行目录>
盘点脚本会生成:
manifest.json:素材清单、原始路径、文件类型、大小、文件类型建议、路径建议产品和图片路径建议分类;summary.md:人类可读统计;- 不修改输入素材,不写入 Obsidian。
若 zip 中文目录乱码,脚本会尝试从 cp437 修复为 gb18030。若仍无法修复,保留原始名并在摘要中提示。
2. 判断文件类型
先按扩展名划分:
- 图片类素材:jpg、jpeg、png、webp、heic、gif、bmp、tif、tiff
- 视频类素材:mp4、mov、m4v、avi、mkv、webm、wmv
- 文档类素材:pdf、doc、docx、xls、xlsx、csv、ppt、pptx、pages、numbers、key、txt、md
无法识别文件类型时,进入根目录 00_待人工确认,不要硬塞进图片、视频或文档。
3. 判断产品
产品通常来自人工标注、原始文件夹名或文件名。优先按路径识别产品;若产品不明确,放到该文件类型下的 00_产品待确认。
示例:
图片类素材/菠萝酒/...
视频类素材/玫瑰酒/...
文档类素材/三华李酒/...
文档类素材/00_产品待确认/...
4. 仅图片类素材判断图片分类
图片类素材允许四个正式图片分类:
01_产品实拍02_好评截图03_用户晒单04_用户咨询
无法达到高置信时使用:
00_待人工确认
判定优先级固定为:
用户咨询 > 好评截图 > 用户晒单 > 产品实拍 > 待人工确认
视频类素材和文档类素材不进入这个优先级,不区分产品实拍、好评截图、用户晒单、用户咨询。
5. 生成入库预案
用脚本生成预案。没有人工决策表时,脚本只用路径高置信规则;人工或 Codex 视觉判断后的决策可写入 CSV 再传入。
python3 scripts/plan_import.py <manifest.json> \
--target-root "九江双蒸知识库v1.0/一、私域库/01_产品库/01.4_产品素材库" \
--output-dir <本次运行目录>
如有人工决策表:
python3 scripts/plan_import.py <manifest.json> \
--target-root "九江双蒸知识库v1.0/一、私域库/01_产品库/01.4_产品素材库" \
--decisions <decisions.csv> \
--output-dir <本次运行目录>
决策表字段:
id,product,category,confidence,note
category 只对图片类素材生效;视频类和文档类会忽略该字段。
预案输出:
import_plan.jsonimport_preview.md
写入前必须把 import_preview.md 给用户确认。没有明确确认时,不运行正式复制。
6. 确认后复制入库
用户确认预案后才执行:
python3 scripts/import_materials.py <import_plan.json> --execute
执行规则:
- 只复制,不移动;
- 不删除原始素材;
- 不覆盖已有文件;
- 同名冲突时自动追加短 hash;
- 统一追加
素材入库总表.md、待人工确认清单.md、异常清单.md。
高风险停止点
遇到以下情况必须停止或放入待人工确认:
- 用户要求删除、移动、覆盖或改名原始素材;
- 目标路径不是九江双蒸项目内的 Obsidian 产品素材库,且用户没有明确确认;
- 文件类型无法判断;
- 产品归属与原始目录明显冲突;
- 图片同时像“好评截图”和“用户咨询”;
- 图片聊天文字看不清;
- 视频或文档被要求继续细分成产品实拍、好评截图、用户晒单、用户咨询;
- 素材可能包含隐私,用户要求直接公开发布;
- 用户要求从素材推断功效、疗效、用户身份、购买动机或合规结论。
输出要求
最终答复要包含:
- 输入素材数量和文件类型;
- 自动入库数量、待人工确认数量、异常数量;
- 每个文件类型和产品的数量;
- 图片类素材的图片分类数量;
- 是否执行了复制;
- 索引文件位置;
- 准确率策略说明:自动分类只对高置信素材执行,低置信素材已进入待人工确认。
不要声称“准确率已达 98%”除非有真实抽检记录支持;没有抽检时只能说“按 98% 目标策略执行”。
What ships with it: 9 files
34.4 KB alongside SKILL.md, 4 of them executable
agents/
- openai.yaml331 B
examples/
references/
- category-rules.md3.1 KB
- product-name-map.md1.1 KB
- target-structure.md2.3 KB
scripts/
- import_materials.pyruns4.7 KB
- inspect_materials.pyruns5.7 KB
- material_common.pyruns7.4 KB
- plan_import.pyruns8.9 KB
Gives 0 of the 12 instructions most product growth skills give in ~2.0k tokens
Counted across 728 of the 1,010 authors here whose files we hold, read 2026-08-07
- Read product marketing context before asking questionsin 24 of 728, across 18 files
- Define the ideal customer profilein 21 of 728, across 3 files
- Document a rollback plan before deploymentin 21 of 728, across 12 files
- Analyze the codebase to understand the productin 19 of 728, across 1 file
- Ask clarifying questions about the value propositionin 19 of 728, across 1 file
- Search for companies matching the criteriain 19 of 728, across 1 file
- Look for signals of immediate needin 19 of 728, across 1 file
- Assign a fit score from one to tenin 19 of 728, across 1 file
- Identify the target decision-maker rolein 19 of 728, across 1 file
- Suggest a personalized contact strategyin 19 of 728, across 1 file
- Provide conversation starters for outreachin 19 of 728, across 1 file
- Format results in a scannable markdown templatein 19 of 728, across 1 file
Said here and by no other author read
- determine file type before product
- categorize only images into user scenarios
- append all imports to the shared master list
- read category rules before classifying
- read target structure before generating paths
- run read-only material inspection first
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.