数据门户架构顾问
安装方式
先用 Prompt 核对来源,也可切换到安装命令或下载可用的安装包。
用 AI 助手安装
复制提示词,让助手先核对固定版本和依赖。
查看 Prompt收起 Prompt
先检查当前 Agent 是否已安装 SkillPI Skill Finder。如果尚未安装,请按官方仓库的说明安装并启用:https://github.com/Findskills/skillpi。然后使用 Skill Finder 打开此 SkillPI 详情页并定位 "portaljs-architect":https://skillpi.com/zh/skills/deepskill-9f6c6a8bbea914a3263c1dd2ba3affd3277eed65986a34e59fb0566b3c29095c。安装前核对上游来源和版本:https://github.com/jeremylongshore/tons-of-skills-marketplace/blob/6ea46b08dce892944db7c16f28d0b5308d402402/plugins/community/portaljs/skills/portaljs-architect/SKILL.md。来源不一致时停止并说明差异。核对通过后,用 Skill Finder 安装目标 Skill,并报告安装路径和来源版本。
此命令从 GitHub 来源仓库获取指定技能。运行前请先查看源文件。
npx skills add https://github.com/jeremylongshore/tons-of-skills-marketplace/tree/6ea46b08dce892944db7c16f28d0b5308d402402/plugins/community/portaljs/skills/portaljs-architect --skill portaljs-architect来源与配置
查看 GitHub 源码jeremylongshore/tons-of-skills-marketplace固定提交 6ea46b08dce892944db7c16f28d0b5308d402402
配置方式待核实
请查看原始来源中的适用工具、依赖和权限要求。
安装实测待补充
本条目介绍 数据门户架构顾问 的公开用途。DeepSkill 当前公开目录说明:“根据既定需求推荐数据门户架构(包括存储、计算、目录、访问、托管和元数据),然后将任务移交至构建技能。作为咨询入口点,适用于启动新的数据门户项目且底层架构尚未确定的场景。” 适用场景包括写代码与代码审查、部署发布与上云。本页根据公开目录资料和固定版本的上游仓库整理,建议使用前核对该版本的环境、工具、权限及数据处理要求。Skillπ 提供站方用途介绍与原站链接;原始 SKILL.md 和附件仅在对应许可允许时开放。
文件浏览
---
name: portaljs-architect
description: Recommend a data-portal architecture (storage, compute, catalog, access, hosting, metadata) from stated needs, then hand off to the build skills. The advisory entry point. Use when starting a new data-portal project and the underlying architecture has not yet been decided.
allowed-tools: Read, Write, Bash(du:*), Bash(wc:*), Bash(ls:*), Bash(head:*)
version: 1.0.0
author: Datopian <hello@datopian.com>
license: MIT
compatibility: Claude Code with PortalJS portals (Next.js 14, React 18, Node 18+). Runs from any project via the plugin, a personal ~/.claude/commands install, or a portaljs clone.
tags:
- portaljs
- data-portal
- architecture
- advisory
- ckan
- nextjs
---
# PortalJS — Architect
## Overview
The advisory entry point for a PortalJS project. Before anything gets scaffolded, this
skill works out what to build: given the kind of portal, the shape of the data, and its
purpose, it fills six architecture slots (storage, catalog, compute, access, hosting,
metadata), resolves two build-time knobs (per-dataset data tier and the portal-wide
`DATA_QUERY` mode), and hands off to the concrete build skills. It decides; it does not
build. When the brief is thin it interviews in short rounds and never dead-ends — every
question has a sensible default, reachable by replying "use defaults."
## Prerequisites
- A rough idea of the portal's purpose and data (exact numbers are not required — the
interview supplies defaults for anything missing).
- Optional: local files or a directory of sample data to inspect for size and shape.
- No PortalJS project needs to exist yet; this skill runs before scaffolding.
## Instructions
The canonical, full step-by-step workflow lives in
[`.claude/commands/portaljs-architect.md`](https://github.com/datopian/portaljs/blob/main/.claude/commands/portaljs-architect.md) —
that file is the single source of truth. Follow it when executing this skill:
1. Parse `$ARGUMENTS` for anything already specified, then interview for what's
missing, one round at a time: (1) what's being built, (2) what the data is,
(3) what it's for, (4) constraints. Accept "use defaults" at any point. Inspect
named files/directories with `du -sh` and line counts to ground size guesses.
2. Derive the recommendation by matching the answers against the decision tables —
Storage/Catalog/Compute by data volume and query needs, Access/Hosting by public
vs. private, Metadata by standards-compliance needs — then resolve the two
build-time knobs: per-dataset **data tier** (`inline | LFS | external`) and the
portal-wide **`DATA_QUERY`** mode (`flat | duckdb`).
3. Echo the architecture brief (stack, reasoning per slot, deviations from default,
deferred items) and wait for confirmation ("go") or corrections.
4. On confirmation, persist the brief to `./ARCHITECTURE.md` in the working directory.
5. Hand off to the build skills — `/portaljs-new-portal`, `/portaljs-add-dataset`,
`/portaljs-connect-ckan`, `/portaljs-define-schema`, `/portaljs-deploy` — mapped
from the brief, and offer to run the first one.
## Output
- **Created:** `./ARCHITECTURE.md` documenting the six slots, the two build-time
knobs, the reasoning, and anything deferred to a later build step.
- **Modified:** nothing else — this skill is advisory only.
- **Verified:** the brief was echoed back and confirmed before being persisted.
- **Result:** a concrete, named sequence of follow-up skill invocations
(e.g. `/portaljs-new-portal` → `/portaljs-add-dataset` → `/portaljs-deploy`).
## Error Handling
| Symptom | Cause | Fix |
| --- | --- | --- |
| Skill keeps asking rounds of questions | Brief was thin or `$ARGUMENTS` omitted | Answer inline, or reply "use defaults" to accept the opinionated default stack |
| Recommendation looks generic | Rounds were skipped without real data details | Give actual size/shape/cadence, or point at files for `du -sh` inspection |
| `ARCHITECTURE.md` never appears | Confirmation step was skipped | Reply "go" once the echoed brief looks right |
| Scaffolded portal has the wrong `DATA_QUERY` | Flat downgrade wasn't applied | Run the `perl -pi -e` one-liner from the command file against `lib/datasets.ts` |
| Hand-off names a skill that doesn't exist | Decision maps to a *(planned)* skill (e.g. `/connect-openmetadata`) | Treat it as designed-in/built-later; proceed with the closest available skill |
## Examples
### Example 1 — National statistics office, DCAT-AP harvesting
```
/portaljs-architect We're a national statistics office. ~200 datasets, mostly large
CSVs (some GBs), updated quarterly, all public, and we must publish DCAT-AP for the
EU data portal.
```
Infers a multi-publisher, analytics-grade portal. Recommends Parquet on R2 + DuckLake +
DuckDB, static Cloudflare Pages, Frictionless + DCAT-AP metadata, `owner` namespace,
data tier `external` for the Parquet, `DATA_QUERY=duckdb`. Writes `ARCHITECTURE.md` and
hands off to `/portaljs-new-portal` then `/portaljs-add-dataset`.
### Example 2 — Small nonprofit, no arguments given
```
/portaljs-architect
```
Runs the full four-round interview since nothing was pre-filled. Accepting defaults at
each round lands on the opinionated default stack: repo files or Git-LFS + R2 storage,
`datasets.json` catalog, DuckDB compute, static access on Cloudflare Pages, Frictionless
metadata, `theme` namespace, data tier LFS, `DATA_QUERY=duckdb`.
### Example 3 — Internal catalog with restricted datasets
```
/portaljs-architect Internal engineering data catalog, single team, dozens of CSVs,
some of it access-controlled to specific roles.
```
The private-data answer in Round 2 flips Access/Hosting to runtime + backend RBAC on
Cloudflare Workers — flagged as the larger, opt-in build — while Storage/Catalog/Compute
still follow the volume-based defaults.
## Resources
- Full workflow: [`.claude/commands/portaljs-architect.md`](https://github.com/datopian/portaljs/blob/main/.claude/commands/portaljs-architect.md)
- Reference: [`references/reference.md`](references/reference.md)
- Decision framework: [`site/content/docs/architecture/decision-framework.md`](https://github.com/datopian/portaljs/blob/main/site/content/docs/architecture/decision-framework.md)
- Related skills: `/portaljs-new-portal`, `/portaljs-add-dataset`, `/portaljs-connect-ckan`, `/portaljs-define-schema`, `/portaljs-deploy`
当前收录版本
1.0.0
质量评测
满分 100。前四项来自 DeepSkill 收录的质量报告。
评分依据
为数据门户架构提供咨询入口点,适用于启动新数据门户项目且底层架构未确定的场景,具有明确的实用价值。
评分依据
触发词精准完整,覆盖多种变体,边界清晰,用户能明确理解何时使用此 Skill。
评分依据
文档结构完整,包含清晰的步骤划分、端到端示例、错误处理机制和资源链接,frontmatter 完整,易于理解和执行。
评分依据
工具来源明确,参数正确,覆盖了主要异常情况,并提供了具体的修复建议,执行流程可靠。
计算方式
取前四项原始分值的算术平均,四舍五入取整。可通过收录报告核对各项评分与说明,运行记录需单独查证。
权限与使用前检查
根据当前主文档自动整理,命中内容可能是示例或防护说明;不是实际权限声明,也未覆盖附件、外部依赖或运行行为。
未从主文档识别到上述常见操作,不能据此判断无需权限或没有风险。
文档引用的域名或地址
这些地址可能用于文档链接或示例,不是已批准的联网清单。使用前核对实际目的地及发送的数据。
- github.com
安全评级与访问范围
risk · 静态证据核验,不代表运行安全保证。
使用前请由宿主核对并授权实际文件读写、外部服务和联网目的地;本站的核验不会授予任何运行权限。
用户讨论
还没有已核实的点评。