Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Write

ASecurity

Guided end-to-end document writing workflow — runs all phases in sequence with approval gates between each

8 stars
0 votes
0 copies
0 views
Added 9/20/2026
documentationgo

Security Analysis

A100/100

Scanned 9/20/2026

Install to Claude Code

$npx -y skills add tstapler/dotfiles --skill write --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Write?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Write
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tstapler-write/badge)](https://www.skillsdirectory.com/skills/tstapler-write)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
description: Guided end-to-end document writing workflow — runs all phases in sequence with approval gates between each
---

# Write a Document (Full Workflow)

Orchestrates the complete document writing workflow: audience definition → structure → draft → review. Each phase produces an artifact; each gate requires approval before proceeding.

**Document topic**: $ARGUMENTS

If no topic provided, ask: "What document are you writing? Give me a working title or topic."

---

## Phase 1: Audience & Purpose

Run this interview before writing a single word. Ask questions one at a time.

### Q1 — Reader role
"Who is the primary reader? What's their role and their knowledge level *relative to this specific topic*?"

### Q2 — Reader goal
"What does the reader need after reading this? One of:
- Complete a specific task
- Learn a skill by doing it
- Look up a specific fact or value
- Understand *why* something works a certain way"

### Q3 — Reader gap
"What does the reader NOT know yet that you do? What's the biggest curse-of-knowledge risk?"

### Q4 — Scope boundary
"What is explicitly out of scope — even if related?"

### Q5 — Type confirmation
Based on answers 1–4, propose the Diataxis type:

| Type | Reader state | Your job |
|---|---|---|
| **Tutorial** | Doesn't know what they don't know | Guide by doing — no explanation mid-task |
| **How-to** | Knows goal, needs the path | Steps only — omit context |
| **Reference** | Needs exact answer | Complete, accurate, scannable |
| **Explanation** | Wants to understand why | Context, tradeoffs, mental models — no instructions |

Say: "This sounds like a **[type]** because [one-line reason]. Does that fit?"

### Phase 1 output — Audience-Purpose Statement
```
Document: [title]
Type: [Diataxis type]

Audience: A [reader role] with [knowledge level relative to this topic]
Goal: reads this to [specific, concrete outcome]
Assumed knowledge: [2–4 items]
Out of scope: [explicit boundary]
```

**Gate 1**: Print the statement. Ask: "Does this accurately capture what you're writing? Any corrections before I build the outline?"

Wait for confirmation. Revise if needed. Do not proceed until approved.

---

## Phase 2: Outline

Generate the skeleton using the confirmed type. Use the correct structural pattern:

**Tutorial**: Learning objective → prerequisites → numbered steps (each does one thing) → verify it works → what's next

**How-to**: Verb-phrase title → prerequisites (only what's required) → numbered imperative steps → result

**Reference**: Entry name → type/signature → one-sentence description → parameters/values → minimal example → constraints. Repeat per entry.

**Explanation**: What this is → the problem it solves → mechanism → adjacent concept comparison → tradeoffs → when to use / when not to → further reading

Fill in real section titles (not placeholders). Add a one-line annotation per section explaining what goes there. Include at the top: `Type: [type] | Estimated length: ~[N] words`

Flag any section that risks type-mixing.

**Gate 2**: Print the outline. Ask: "Does this structure match what you need? Any sections to add, remove, or reorder?"

Wait for confirmation. Revise if needed. Do not proceed until approved.

---

## Phase 3: Draft

Write the complete document from the confirmed outline. Apply all rules below without exception.

### Type discipline
- **Tutorial**: Steps produce visible, checkable state changes. Never explain *why* mid-task.
- **How-to**: Steps only. Zero background. If understanding is required, it's a prerequisite, not inline prose.
- **Reference**: Structure over narrative. Every entry identical shape. Complete and accurate above all else.
- **Explanation**: No instructions. Builds mental model only. Does not tell the reader to do anything.

### Every Page is Page One
- First paragraph establishes context — assume the reader arrived directly, not sequentially
- Serve one purpose; link to adjacent topics rather than containing them
- End with navigation: where to go next

### Compression
- Lead with the most important thing
- Start sentences with verbs where possible
- Cut without mercy: "you can", "there is/are", "in order to", "it should be noted", "basically", "simply", "just", "very"
- One idea per sentence
- Concrete example over abstract description

### Three Questions test — apply before delivering
1. What specific decision or action does the reader need?
2. What prevents them from acting confidently?
3. Does every paragraph serve question 2? If not, cut it.

### Quality gate (self-check before outputting)
- [ ] Every section title matches the declared type's pattern
- [ ] No section mixes types
- [ ] First sentence states the outcome or purpose — not context
- [ ] Reader can complete the stated goal without a follow-up question the document could have answered

After the document, append a **Compression log** (2–4 bullets: what was cut and why).

**Gate 3**: Print the draft. Ask: "Does this draft work? Options:
1. ✅ Done — I'll take it as-is
2. 🔍 Run clarity review (`/docs:review-clarity`)
3. ✏️ Polish the writing (`/docs:refine-writing`)
4. 🔄 Revise — [describe what to change]"

---

## Phase 4: Review (if requested)

If the user selects option 2 (clarity review), apply the full Three Questions Framework + cognitive load review from `/docs:review-clarity`.

If the user selects option 3 (polish), apply compression and style rules from `/docs:refine-writing`.

Deliver the revised document. Return to Gate 3.

---

## Saving Artifacts

At any gate, if the user says "save this", write:
- Audience-purpose statement → `doc_plans/<topic>/audience-purpose.md`
- Outline → `doc_plans/<topic>/outline.md`
- Draft → `doc_plans/<topic>/draft.md`
- Final → `doc_plans/<topic>/<topic>.md`

Attribution

tstaplertstapler
View sourceMore from tstapler →
SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

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 (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

Related Skills

Context Fundamentals

Understand the components, mechanics, and constraints of context in agent systems. Use when designing agent architectures, debugging context-related failures, or optimizing context usage.

179001 votes

release-notes

Draft release notes and changelog entries from git history or merged PRs between two refs (tags/SHAs/branches), including breaking changes, migrations, and upgrade steps. Use when the user asks for release notes, changelog updates, or a GitHub Release draft.

1301 votes

docs-style-guide

Documentation style guide enforcer by @planetabhi. Applies and reviews the writing style guide when authoring or editing product documentation and tutorials. Use to check prose for voice, tense, word choice, inclusive language, formatting, code block, UI, Markdown, and number/date conventions.

11 votes

Caveman Help

Quick-reference card for all caveman modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /caveman-help, "caveman help", "what caveman commands", "how do I use caveman".

1023330 votes

How It Works

Explain how claude-mem captures observations, when memory injection kicks in, and where data lives. Use when the user asks "how does claude-mem work?" or "what is this thing doing?".

942310 votes
View all in documentation →