Skip to content
Back to skills

Julien Ref Doc Production

ASecurity

Documentation production methodology with DRY principles. Covers 8 document types (GUIDE, INCIDENT, SESSION, AUDIT, RECHERCHE, etc.), naming conventions, and YAML metadata.

  • 666 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added February 9, 2026
documentationsecurityperformancedocumentation

Security analysis

A100/100

Scanned February 12, 2026

npx -y skills add majiayu000/claude-skill-registry --skill julien-ref-doc-production --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Julien Ref Doc Production?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Julien Ref Doc Production
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-julien-ref-doc-production/badge)](https://www.skillsdirectory.com/skills/majiayu000-julien-ref-doc-production)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: julien-ref-doc-production
description: Documentation production methodology with DRY principles. Covers 8 document types (GUIDE, INCIDENT, SESSION, AUDIT, RECHERCHE, etc.), naming conventions, and YAML metadata.
triggers:
  - comment documenter mon projet
  - aide-moi à structurer ma doc
  - je veux écrire la doc
  - modèle de document
  - écrire un rapport
  - how to document my project
  - help me write documentation
  - project documentation guide
  - create documentation
  - incident report
  - session report
  - guide template
---

# Documentation Production Methodology

> **Version**: 2.0.0-DRY

---

## Principles

1. **DRY**: Avoid duplication, use cross-references
2. **Living docs**: Update rather than create, archive obsolete
3. **Targeted audience**: Humans | LLM | All + level (1-10)
4. **Actionable**: Concrete examples, reproducible commands

---

## Structure

```
project/
├── docs/
│   ├── guides/        # Evolving docs
│   └── rapports/      # Event-based docs
└── README.md          # Index
```

---

## Naming Conventions

### EVOLVING Docs (living docs)
**Types**: GUIDE, METHODOLOGIE, AUDIT, TROUBLESHOOTING
**Name**: `[TYPE]_[subject].md` (no date)
**Timestamp**: In content, per section

**Required YAML header**:
```yaml
---
title: "Guide X"
version: X.Y.Z
created: YYYY-MM-DD
updated: YYYY-MM-DD
authors: [...]
audience: ["Humans"|"LLM"|"All"]
level: "Beginner (1-3)" | "Intermediate (4-6)" | "Expert (7-10)"
status: "Active" | "Archived" | "Draft"
tags: [...]
related: [related files]
---
```

### EVENT-BASED Docs (snapshots)
**Types**: INCIDENT, SESSION, RECHERCHE, RECOMMANDATIONS, MIGRATION, PERFORMANCE, SECURITE
**Name**: `YYYY.MM.DD_HH.mm_[commit-short]_[TYPE]_[subject].md`
**Timestamp format**: `2025.10.14_12.33` (dots + underscores)

**Required metadata**:
```markdown
## Metadata
- **Date**: YYYY-MM-DD HH:MM
- **Commit**: abc1234 (7 chars)
- **Duration**: X hours (if applicable)
```

---

## Document Types

### INCIDENT (event-based)
**When**: Blocking problem resolved

```markdown
# Incident: [Title]

## Metadata
- Date, Duration, Severity, Commit

## Summary
[2-3 sentences]

## Symptoms → Timeline → Cause → Solution → Prevention

## References
```

---

### SESSION (event-based)
**When**: Significant session (feature, refactor)

```markdown
# Session: [Title]

## Metadata
- Date, Duration, Commits

## Objective → Achievements → Decisions → Problems → Modified files → Tests → Next

## References
```

---

### GUIDE (evolving)
**When**: Reusable procedure

```markdown
---
[Complete YAML header]
---

# Guide: [Title]

## Overview → Concepts → Step-by-step → Troubleshooting → Examples → References
```

---

### AUDIT (evolving or event-based)
**When**: Pre-intervention analysis

```markdown
# Audit: [Title]

## Metadata

## Executive Summary
- Score, Risks, Recommendations

## Findings (Critical, Major, Minor)

## Prioritized recommendations → Action plan → References
```

---

### RECHERCHE (event-based)
**When**: Investigation, exploration, benchmark

```markdown
# Research: [Title]

## Metadata
- Date, Objective, Duration

## Question → Methodology → Discoveries → Options → Comparison → Recommendation → References
```

---

### RECOMMANDATIONS (event-based)
**When**: Important technical decision

```markdown
# Recommendations: [Title]

## Metadata

## Context → Evaluated options (Pros/Cons/Evaluation) → Decision matrix → Recommendation → Risks → Next
```

---

### MIGRATION (event-based)
**When**: Tech/version/infra migration

```markdown
# Migration: [Title]

## Metadata

## Objective → Before/After → Procedure → Verifications → Problems → Rollback plan → Lessons → Follow-up
```

---

### PERFORMANCE (event-based)
**When**: Measurable optimization

```markdown
# Performance: [Title]

## Metadata

## Summary → Metrics before → Optimizations → Metrics after → Charts → Recommendations → References
```

---

### SECURITE (event-based)
**When**: Security audit, vulnerabilities

```markdown
# Security Audit: [Title]

## Metadata

## Summary → Vulnerabilities (Critical/Major/Minor) → Best practices → Remediation plan → Verifications → References
```

---

## Workflow

**Create report when**:
- Session > 1h with significant changes → SESSION
- Blocking problem resolved → INCIDENT
- Investigation completed → RECHERCHE
- Technical decision → RECOMMANDATIONS
- New procedure → GUIDE (or update)
- Migration completed → MIGRATION
- Measurable optimization → PERFORMANCE
- Vulnerability found → SECURITE

**Production**:
1. Choose type
2. Use template
3. Fill all sections
4. Name correctly
5. Add metadata
6. Reference in README
7. Commit

---

## Quality Checklist

### Content
- [ ] All sections filled
- [ ] No TODO/To complete
- [ ] Commands tested
- [ ] Paths correct
- [ ] Links valid

### Form
- [ ] Naming respected
- [ ] Metadata present
- [ ] Valid Markdown
- [ ] Code blocks with language

### Context
- [ ] Audience identified
- [ ] Level adapted
- [ ] Cross-references

### Traceability
- [ ] Commits mentioned
- [ ] Files listed
- [ ] Timeline clear

---

## Related Skills

- **julien-ref-doc-review**: Use after writing to annotate and review documentation with @FIX, @TODO, @VERIFY tags

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.