duckdb-analyst
Use for any analytical question over ./data/*.parquet. Prefer DuckDB, always show the SQL.
Best use case
duckdb-analyst is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Use for any analytical question over ./data/*.parquet. Prefer DuckDB, always show the SQL.
Teams using duckdb-analyst should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/duckdb-analyst/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How duckdb-analyst Compares
| Feature / Agent | duckdb-analyst | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Use for any analytical question over ./data/*.parquet. Prefer DuckDB, always show the SQL.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# DuckDB Analyst
## Conventions
- Data lives in `./data/*.parquet` (a read-only mount). Never write there.
- Use DuckDB — either the `duckdb` CLI or `python -c "import duckdb; …"`.
- **Always print the SQL** you ran — analysts trust answers only when they see the query.
- Cap queries to `LIMIT 1000` by default.
- Save intermediate results as `./cache-<slug>.parquet` in the workdir.
## Workflow
1. `ls ./data` to discover files
2. `duckdb -c "DESCRIBE SELECT * FROM read_parquet('./data/X.parquet') LIMIT 0"` to learn schema
3. Write the query and run it via `duckdb -c "..."` or a short Python script
4. If the user wants a chart, follow the `matplotlib-charts` skill next
## Guardrails
- If the underlying file is larger than ~10 GB, warn the user before scanning it
- Never suggest or run `DELETE` / `UPDATE` / `DROP` — DuckDB on parquet can't anyway, but the LLM must still not suggest it
## Example shell
```bash
duckdb -c "SELECT product, SUM(revenue) AS total
FROM read_parquet('./data/sales.parquet')
GROUP BY product ORDER BY total DESC LIMIT 1000"
```Related Skills
skill-creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
support-triage
Use for every inbound support ticket. Defines triage policy, tone, and escalation rules.
zootopia-ppt
将用户的演示讲稿转化为拟人化动物世界主题的演示文稿图像（顶级3D动画电影风格）。 分三步：① 理解演讲内容与受众 ② 生成每页的3D动画电影风格生图提示词（Markdown 格式，保存到 workspace/docs/outline/） ③ 调用 image_gen_ppt.py 批量生成图像（保存到 workspace/docs/slides/）。 触发条件：用户提供演示稿/PPT 内容，要求制作 疯狂动物城风格 / 拟人化动物 / 3D动画电影风格的图像或演示文稿。
pro-ppt
将用户的演示讲稿转化为「专业 · 高级感」风格的演示文稿图像（高端商务编辑风）。 视觉特征：极浅灰主色（#F2F2F2）、金属灰辅色、金色点缀、深蓝灰背景（#0F233B），整体克制、克莱因蓝调、印刷品级质感。 分三步：① 理解演讲内容与受众 ② 生成每页高级商务视觉提示词（Markdown 格式，保存到 workspace/docs/outline/） ③ 调用 image_gen_ppt.py 批量生成图像（保存到 workspace/docs/slides/）。 触发条件：用户提供演讲稿/PPT 内容，要求制作 专业 / 高级感 / 高端商务 / 企业级 / 资本市场 / 咨询风 / 编辑风 的演示文稿。
ocr
OCR image files using the Qwen VL model via the bundled scripts/ocr.py (lives in this skill's directory, NOT your cwd). Use when the user asks to extract text from an image, perform OCR on a photo or screenshot, or recognize characters in an image file (.jpg, .jpeg, .png, .gif, .webp). Requires QWEN_API_KEY and QWEN_BASE_URL (env vars, or a .env in the skill directory).
kanban-tasks
使用 agentao-kanban 看板进行任务管理（拆任务 / 状态流转 / 后台服务 / 看产物）。 本技能负责：① 初始化 .kanban/ 项目 ② 把需求拆成原子卡片入板 ③ 推动状态流转 INBOX → READY → DOING → REVIEW → DONE ④ 启停 daemon 与 Web UI ⑤ 察看卡片状态与产物。 触发关键词（中/英）：看板 / kanban / 任务管理 / task board / 拆任务 / 任务卡 / 排期 / WIP / dispatcher / orchestrator。 复合触发短语（一键启停）： • "启动 kanban" → daemon（mock 执行器，后台）+ Web UI（自动开浏览器） • "启动 kanban 真跑" → daemon（--executor agentao，真调 sub-agent）+ Web UI • "停止 kanban" → 先停 Web UI，再停 daemon
matplotlib-charts
Produce PNG charts with matplotlib. Save to chart-<slug>.png in the working directory.
daily-digest
Build a daily digest from curated sources. Follow the output contract strictly.
fswatch-duckdb
FileSystemWatcher over /tmp with DuckDB/DuckLake persistence. Auto-starts on Amp sessions for resilient file monitoring with temporal queries.
duckdb-timetravel
Layer 3: Temporal Versioning and ACSet Schema Generation for DuckDB
duckdb-quadruple-interleave
Chaotic interleaving across local DuckDB databases modeled as coupled quadruple pendula. Random walks both BETWEEN databases and WITHIN tables for context injection.
duckdb-ies
Layer 4: IES Interactome Analytics with GF(3) Momentum Tracking