writing:draft

A drafting command that turns an outline and its research into prose while applying selected style and voice guidance.

In plain words
What is it for?
Use it to draft blog posts, newsletters, social content, or documentation section by section with review checkpoints.
Why use it?
It removes the need to build a full draft manually from scattered outline, source, and style files.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/robertguss/claude-code-toolkit/writing-draft
Clone the repo
git clone --depth 1 https://github.com/robertguss/claude-code-toolkit
Per session 12 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,159 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00012 $0.01159
Opus 5 $0.00006 $0.00580
Sonnet 5 $0.00002 $0.00232
Haiku 4.5 $0.00001 $0.00116

Measured 2d ago against content hash a90f1485f765, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

writing:draft scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/compound-writing/commands/writing-draft.md · 198 lines

How it starts

The opening of the file, as written. The whole thing — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Writing Draft Command

Execute an outline into prose, following style preferences and voice profiles.

Input

<outline_path> #$ARGUMENTS </outline_path>

Workflow Overview

This command executes the drafting phase:

  1. Load outline and research
  2. Load applicable style guide
  3. Pre-draft checklist
  4. Section-by-section drafting with voice guardian loop
  5. Quality checkpoints

Phase 1: Load Context

Read Outline and Research

Read the outline file at [outline_path]
Read associated research.md and sources.md in same directory
Extract:
- The thesis/argument
- The hook
- Section structure
- Source requirements per section

Load Style Guide

Check for applicable style:

  1. Check for voice profile: .claude/voice-profiles/[name].yaml
  2. Check for style skill: every-style-editor, dhh-writing, pragmatic-writing
  3. Ask if none found: "Which style should I use? (Every style, DHH style, Pragmatic style, or provide custom)"

Extract Channel Guidance

Based on outline metadata or ask:

  • Blog: longer form, storytelling allowed
  • Newsletter: personality forward, direct address okay
  • Social: punchy, hooks required
  • Documentation: clear, step-by-step, minimal personality

Phase 2: Pre-Draft Checklist

Before writing any prose, verify:

  • Opening is concrete - Story or example first, not theory
  • Each section has clear purpose - Know what each must accomplish
  • Sources are sufficient - Every claim can be supported
  • Voice is defined - Know the style to match

If any fail, return to /writing:plan to address.

Phase 3: Section-by-Section Drafting

The Producer-Critic Loop

For each section, use iterative voice-guardian feedback:

Loop for each section:
  1. Draft the section following:
     - Outline beats
     - Style guide rules
     - Baseline strategies (short sentences, active voice, concrete examples)

  2. Run voice-guardian check:
     Task voice-guardian: "Check this section against voice profile:
     [section text]
     Voice profile: [profile details]
     Return score (0-100) and specific fixes needed."

  3. If score < 85:
     - Apply suggested fixes
     - Re-check until score ≥ 85

  4. If score ≥ 85:
     - Move to next section

Read the full file on GitHub · 198 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 198 lines · 12 tokens per session scan A a90f1485f765

Subscribe to this mod's changes

writing:draft is a command published in the GitHub repository robertguss/claude-code-toolkit (108 stars, last pushed 24d ago), licensed MIT. It adds 12 tokens to every session and 1,159 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.