菜单
Run read-only SQL against the warehouse through the Redshift Data API helper, with masking, timeouts, and row caps applied. Use for every data question.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/PacktPublishing/Developing-AI-ML-Solutions-on-AWS --skill redshift-data-api命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | redshift-data-api |
| description | Run read-only SQL against the warehouse through the Redshift Data API helper, with masking, timeouts, and row caps applied. Use for every data question. |
Run queries through helper.py, never through a direct database connection.
The default way is the CLI, one command, no Python of your own:
uv run skills/redshift-data-api/helper.py \
"SELECT state, AVG(score) AS avg_score
FROM analytics.applicant_credit_profile
GROUP BY state ORDER BY avg_score DESC LIMIT 10"
Each result row prints as one JSON object; read the values and present them
as a markdown table. Only import execute_redshift_query in Python when you
genuinely need computation over the rows, and remember rows are dicts keyed
by column name, never tuples.
config/pii_columns.yaml before results reach you.knowledge/catalog.md; check it before
guessing table or column names.