Agent skill
writing-skills
Quellcode ansehen: nickcrew/claude-cortex
Installation
npx skills add nickcrew/claude-cortex --skill writing-skills48
Installationen
EU-hosted inference API
Power your AI agent skills with open-source models.
Drop-in OpenAI-compatible API. No data leaves Europe.
MiniMax
MiniMax M3
$0.50 / $1.90
per M tokens
Z.ai
GLM 5.3 Flash
$0.20 / $0.60
per M tokens
MoonshotAI
Kimi K3
$4.00 / $18.00
per M tokens
DeepSeek
DeepSeek V4.1 Flash
$0.40 / $1.42
per M tokens
Writing Skills
Overview
Writing skills IS Test-Driven Development applied to process documentation.
You write test cases (pressure scenarios with subagents), watch them fail (baseline behavior), write the skill (documentation), watch tests pass (agents comply), and refactor (close loopholes).
Core principle: If you didn't watch an agent fail without the skill, you don't know if the skill teaches the right thing.
Personal skills live in agent-specific directories: ~/.claude/skills for Claude Code, ~/.codex/skills for Codex
REQUIRED BACKGROUND: This skill applies a test-first discipline to skill-writing — watch a skill fail before you trust it. The testing mechanics live in testing-skills-with-subagents (see below).
When to Use
Create a skill when:
- Technique wasn't intuitively obvious to you
- You'd reference this again across projects
- Pattern applies broadly (not project-specific)
- Others would benefit
Don't create for:
- One-off solutions
- Standard practices well-documented elsewhere
- Project-specific conventions (put in CLAUDE.md)
Skill Types
- Technique: Concrete method with steps (condition-based-waiting, root-cause-tracing)
- Pattern: Way of thinking about problems (flatten-with-flags, test-invariants)
- Reference: API docs, syntax guides, tool documentation (office docs)
Quick Reference
| Task | Load reference |
|---|---|
| Understand TDD mapping for skills | skills/writing-skills/references/tdd-mapping.md |
| Learn skill structure and organization | skills/writing-skills/references/skill-structure.md |
| Optimize for search and discovery | skills/writing-skills/references/search-optimization.md |
| Test skills with subagents | skills/writing-skills/references/testing-skills.md |
| Code examples and flowcharts | skills/writing-skills/references/code-and-flowcharts.md |
| Complete creation checklist | skills/writing-skills/references/checklist.md |
RED-GREEN-REFACTOR Workflow
RED: Write Failing Test (Baseline)
- Create pressure scenarios (3+ combined pressures for discipline skills)
- Run scenarios WITHOUT skill - document baseline behavior verbatim
- Identify patterns in rationalizations/failures
This is critical: You must see what agents naturally do before writing the skill.
GREEN: Write Minimal Skill
- Write skill addressing specific baseline failures
- Follow structure guidelines (see skill-structure.md)
- Optimize for search (see search-optimization.md)
- Run same scenarios WITH skill - verify agents now comply
REFACTOR: Close Loopholes
- Identify NEW rationalizations from testing
- Add explicit counters for discipline skills
- Build rationalization table from all iterations
- Re-test until bulletproof
REQUIRED SUB-SKILL: Use testing-skills-with-subagents for complete testing methodology.
The Iron Law
NO SKILL WITHOUT A FAILING TEST FIRSTThis applies to NEW skills AND EDITS to existing skills.
Write skill before testing? Delete it. Start over.
No exceptions:
- Not for "simple additions"
- Not for "just adding a section"
- Not for "documentation updates"
- Delete means delete
Common Mistakes
| Mistake | Fix |
|---|---|
| Writing skill before testing | Delete skill, run baseline test first |
| Skipping baseline (RED phase) | You don't know what to teach without seeing failure |
| Testing with skill already present | Remove skill, get true baseline behavior |
| "Batching" multiple skills | Complete RED-GREEN-REFACTOR for each skill before moving on |
| Vague descriptions | Start with "Use when...", include specific triggers |
| Using @ links to reference skills | Use skill names only, avoid force-loading |
| Multiple mediocre code examples | One excellent example in most relevant language |
| Narrative storytelling | Focus on reusable patterns, not one-off stories |
STOP Before Moving to Next Skill
After writing ANY skill, you MUST STOP and complete the deployment process.
Do NOT create multiple skills in batch without testing each. Deploying untested skills = deploying untested code.
See checklist.md for full deployment checklist.
Resources
- Official guidance: For Anthropic's official skill authoring best practices, see anthropic-best-practices.md
- Graphviz conventions: See @graphviz-conventions.dot for flowchart style rules
- Testing methodology: Use testing-skills-with-subagents
Verwandte Skills
Mehr aus dieser Quelle: nickcrew/claude-cortex
react-performance-optimization
nickcrew/claude-cortex
1.9K installs
owasp-top-10
nickcrew/claude-cortex
760 installs
ui-design-aesthetics
nickcrew/claude-cortex
362 installs
threat-modeling-techniques
nickcrew/claude-cortex
357 installs
token-efficiency
nickcrew/claude-cortex
245 installs
repo-cleanup
nickcrew/claude-cortex
183 installs
So verwenden Sie diesen Skill
- 1
Install writing-skills by running
npx skills add nickcrew/claude-cortex --skill writing-skillsin your project directory. Führen Sie den obigen Installationsbefehl in Ihrem Projektverzeichnis aus. Die Skill-Datei wird von GitHub heruntergeladen und in Ihrem Projekt platziert. - 2
Keine Konfiguration erforderlich. Ihr KI-Agent (Claude Code, Cursor, Windsurf usw.) erkennt installierte Skills automatisch und nutzt sie als Kontext bei der Code-Generierung.
- 3
Der Skill verbessert das Verständnis Ihres Agenten für writing-skills, und hilft ihm, etablierte Muster zu befolgen, häufige Fehler zu vermeiden und produktionsreifen Code zu erzeugen.
Was Sie erhalten
Skills sind Klartext-Anweisungsdateien — kein ausführbarer Code. Sie kodieren Expertenwissen über Frameworks, Sprachen oder Tools, das Ihr KI-Agent liest, um seine Ausgabe zu verbessern. Das bedeutet null Laufzeit-Overhead, keine Abhängigkeitskonflikte und volle Transparenz: Sie können jede Anweisung vor der Installation lesen und prüfen.
Kompatibilität
Dieser Skill funktioniert mit jedem KI-Coding-Agenten, der das skills.sh-Format unterstützt, einschließlich Claude Code (Anthropic), Cursor, Windsurf, Cline, Aider und anderen Tools, die projektbezogene Kontextdateien lesen. Skills sind auf Transportebene framework-agnostisch — der Inhalt bestimmt, für welche Sprache oder welches Framework er gilt.
Data sourced from the skills.sh registry and GitHub. Install counts and security audits are updated regularly.