
A structured, AI-native framework for building scalable content systems.
概览
A skill-based system for content creation. Each skill is a structured package with defined inputs, workflows, templates, and quality criteria. - – Error messages, empty states, dialogs, notifications, forms, onboarding - – UI copy systems, patterns, audits, naming, cognitive load - – Voice/tone, taxonomy, messaging hierarchy, roadmaps - – Feature context packs, vocabulary, state maps, user journeys - – How-to guides, API docs, reference docs, release notes - – CLI help, SDK docs, code comments, configuration guides - – Landing pages, product messaging, campaigns, app store listings - – Blog posts, case studies, thought leadership, tutorials - – Agent instructions, system prompts, tool descriptions, evals - – User research synthesis, feedback analysis, audit reports 1. – Check governance/skill-catalog.md or match by trigger keywords 2. – Understand required inputs and workflow 3. – All outputs follow /shared guidelines 4. – Follow TEMPLATES.md for output format 5.
README
以下为原文 README 节选,完整内容请查看原文。
Hoover Content System (HCS)
A skill-based system for content creation. Each skill is a structured package with defined inputs, workflows, templates, and quality criteria.
What This Is
HCS provides repeatable skills for:
- UX Writing – Error messages, empty states, dialogs, notifications, forms, onboarding
- Content Design – UI copy systems, patterns, audits, naming, cognitive load
- Content Strategy – Voice/tone, taxonomy, messaging hierarchy, roadmaps
- Content Context – Feature context packs, vocabulary, state maps, user journeys
- Technical Documentation – How-to guides, API docs, reference docs, release notes
- Developer Content – CLI help, SDK docs, code comments, configuration guides
- Marketing – Landing pages, product messaging, campaigns, app store listings
- Editorial – Blog posts, case studies, thought leadership, tutorials
- MCP & Agents – Agent instructions, system prompts, tool descriptions, evals
- Research & Insights – User research synthesis, feedback analysis, audit reports
Repo Structure
/skills/ → Skill packages organized by category
/shared/ → Global standards (apply to ALL outputs)
/governance/ → Rules, catalog, glossary, naming conventions
/evals/ → Test cases for skill validation
Skill Package Structure
Each skill folder contains:
| File | Purpose |
|---|---|
SKILL.md |
Entry point. Triggers, inputs, workflow, degrees of freedom. |
TEMPLATES.md |
Output structures and formats. |
RUBRIC.md |
Pass/fail quality criteria. |
EXAMPLES.md |
Input/output pairs showing correct execution. |
How to Use
- Find the skill – Check
governance/skill-catalog.mdor match by trigger keywords - Read SKILL.md – Understand required inputs and workflow
- Apply shared standards – All outputs follow
/sharedguidelines - Use templates – Follow
TEMPLATES.mdfor output format - Validate – Check against
RUBRIC.mdcriteria
Getting Started: Write an Error Message
Here’s a quick walkthrough using the writing-error-messages skill.
1. Gather inputs
You need:
- Error type: validation, system, network, permission, or not-found
- Severity: blocking, degraded, or informational
- User action: what the user was trying to do
- Recovery actions: what they can do next
Example scenario: User tries to save a file but their session expired.
2. Open the skill
Read skills/ux-writing/writing-error-messages/SKILL.md for the workflow.
3. Apply the template
From TEMPLATES.md, error messages use this structure:
Title: [What happened – max 60 chars]
Body: [Why it happened – max 150 chars, optional]
Action: [What to do – max 25 chars]
4. Write the copy
Title: Your session expired
Body: Sign in again to save your changes.
Action: Sign in
5. Validate against the rubric
Check RUBRIC.md:
- ✅ Title under 60 characters
推荐工具
换一个关键词,或者移除筛选条件。
安装
npx skillfish add jeremydhoover-blip/hoover-content-system