list-enrichment

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Table Enrichment

表格Enrichment

Add research-powered enrichment columns to Extruct company tables.

为Extruct企业表格添加由研究驱动的 enrichment 列。

Extruct API Operations

Extruct API 操作

This skill delegates all Extruct API calls to the

extruct-api

skill.

For all Extruct API operations, read and follow the instructions in

skills/extruct-api/SKILL.md

All table reads, column creation, enrichment runs, polling, and data fetching are handled by the extruct-api skill. This skill focuses on what to enrich and how to design columns — the extruct-api skill handles the API execution.

本Skill将所有Extruct API调用委托给

extruct-api

Skill。

对于所有Extruct API操作，请阅读并遵循

skills/extruct-api/SKILL.md

中的说明。

所有表格读取、列创建、enrichment运行、轮询和数据获取均由extruct-api Skill处理。本Skill专注于要增强哪些内容以及如何设计列——extruct-api Skill负责API执行。

Workflow

工作流

1. Confirm the table

1. 确认表格

Get the table ID from the user (URL or ID). Use the extruct-api skill to fetch table metadata. Show the user: table name, row count, existing columns.

向用户获取表格ID（URL或ID）。使用extruct-api Skill获取表格元数据。向用户展示：表格名称、行数、现有列。

2. Get column configs

2. 获取列配置

Two paths:

Path A: From enrichment-design — User has

column_configs

ready. Confirm and proceed.

Path B: Design on the fly — Confirm with the user:

What data point? — what to research (e.g. "funding stage", "primary vertical", "tech stack")
Output format — pick the right format:

Format	When to use	Extra params
`text`	Free-form research output	—
`number` / `money`	Numeric data (revenue, headcount)	—
`select`	Single choice from known categories	`labels: [...]`
`multiselect`	Multiple tags from known categories	`labels: [...]`
`json`	Structured multi-field data	`output_schema: {...}`
`grade`	1-5 score	—
`label`	Single tag from list	`labels: [...]`
`date`	Date values	—
`url` / `email` / `phone`	Contact info	—

Agent type — default
```
research_pro
```
. Use
```
llm
```
when no web research needed (classification from existing profile data).

有两种路径：

路径A：来自enrichment-design — 用户已准备好

column_configs

。确认后继续。

路径B：即时设计 — 与用户确认以下内容：

要添加的数据点？ — 需要研究的内容（如“融资阶段”、“核心行业领域”、“技术栈”）
输出格式 — 选择合适的格式：

格式	适用场景	额外参数
`text`	自由格式的研究输出	—
`number` / `money`	数值型数据（收入、员工人数）	—
`select`	从已知分类中单选	`labels: [...]`
`multiselect`	从已知分类中多选标签	`labels: [...]`
`json`	结构化多字段数据	`output_schema: {...}`
`grade`	1-5分评分	—
`label`	从列表中选择单个标签	`labels: [...]`
`date`	日期值	—
`url` / `email` / `phone`	联系信息	—

Agent类型 — 默认使用
```
research_pro
```
。当无需网页研究时（从现有资料数据进行分类），使用
```
llm
```
。

3. Write the prompt

3. 编写提示词

Craft a clear prompt using

{input}

for the row's domain value. Prompt guidelines:

Be specific about what to find
Specify the exact output format in the prompt (e.g. "Return ONLY a number in millions USD")
Include fallback instruction (e.g. "If not found, return N/A")
For
```
select
```
/
```
multiselect
```
, the labels constrain the output — the prompt should guide which label to pick

使用

{input}

指代行的域名值，编写清晰的提示词。提示词指南：

明确说明需要查找的内容
在提示词中指定确切的输出格式（例如：“仅返回以百万美元为单位的数字”）
包含回退说明（例如：“若未找到，返回N/A”）
对于
```
select
```
/
```
multiselect
```
，标签会限制输出——提示词应指导选择哪个标签

4. Create the column(s)

4. 创建列

Delegate column creation to the extruct-api skill with the

column_configs

array.

将列创建任务委托给extruct-api Skill，传入

column_configs

数组。

5. Trigger enrichment (only the new columns)

5. 触发增强（仅针对新列）

Delegate the enrichment run to the extruct-api skill. Always scope the run to the newly created column(s) only. Avoid broad or implicit run payloads when you only intend to enrich specific columns.

将增强运行任务委托给extruct-api Skill。始终将运行范围限定为仅新创建的列。当仅打算增强特定列时，避免使用宽泛或隐含的运行负载。

6. Monitor progress

6. 监控进度

Delegate progress monitoring to the extruct-api skill. Use it to poll table data and check cell statuses.

Show the user:

Current % complete (done cells / total cells)
Number of failed cells (if any)
Estimated time remaining (based on rate so far)

Stop polling when all cells are done or failed.

将进度监控任务委托给extruct-api Skill。使用该Skill轮询表格数据并检查单元格状态。

向用户展示：

当前完成百分比（已完成单元格数/总单元格数）
失败单元格数量（如有）
预计剩余时间（基于当前速率）

当所有单元格完成或失败时停止轮询。

7. Quality spot-check

7. 质量抽查

After enrichment completes (or after 50%+ is done), fetch a sample of 5-10 enriched rows and display for review.

Present to user as a table. Ask:

"Does the data quality look right?"
"Any columns returning garbage or N/A too often?"
"Should we adjust any prompts and re-run?"

If quality issues are found:

Delete the problematic column
Adjust the prompt
Re-create and re-run

增强完成后（或完成50%以上时），提取5-10条已增强的行样本并展示供审核。

以表格形式呈现给用户。询问：

“数据质量是否符合预期？”
“是否有列频繁返回无效内容或N/A？”
“是否需要调整提示词并重新运行？”

若发现质量问题：

删除有问题的列
调整提示词
重新创建列并重新运行