writer

writer is an agent for coding agents from queelius/claude-anvil. It costs 368 tokens per session (1,407 once invoked), scanned A, original, MIT.

A multi-agent writing workflow for drafting technical non-fiction from a detailed chapter plan.

In plain words
What is it for?
Use it to draft prose sections, worked notebooks, or source-based rewrites, then check and fix the results against the chapter specification.
Why use it?
It coordinates specialist writers and automated audits, helping the draft stay aligned with the plan's content, length, labels, notation, and reference requirements.

Agent

Part of the bookwright plugin — 3 skills, 11 commands, 11 agents, 1 hook shipped together

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 agents/queelius/claude-anvil/writer
Clone the repo
git clone --depth 1 https://github.com/queelius/claude-anvil

Or install bookwright, the plugin that ships this one along with the rest of its 3 skills, 11 commands, 11 agents, 1 hook.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for writer

README.md
[![agentmods](https://agentmods.dev/badge/agents/queelius/claude-anvil/writer.svg)](https://agentmods.dev/agents/queelius/claude-anvil/writer)
Your own site
<a href="https://agentmods.dev/agents/queelius/claude-anvil/writer"><img src="https://agentmods.dev/badge/agents/queelius/claude-anvil/writer.svg" alt="Measured on agentmods" height="20"></a>
Per session 368 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,407 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 $0.00368 $0.01407
Opus 5 $0.00184 $0.00704
Sonnet 5 $0.00074 $0.00281
Haiku 4.5 $0.00037 $0.00141

Measured 3d ago against content hash bc8a46d344c8, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

writer 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 3d 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.

bookwright/agents/writer.md · 106 lines

How it starts

The opening of the file, as written. The whole thing — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You orchestrate multi-agent drafting for technical non-fiction textbooks. You are the lead author: you understand the chapter plan, assign work to the right specialist per section type, audit the results, and loop on fixes until the output meets the plan's standards.

Orchestrated Drafting

Every section goes through: specialist draft, then spec-auditor and quality-auditor in parallel, then a fix loop if needed.

Review-driven fixes are NOT this agent's job: when the input is a review report rather than a chapter plan, the bookwright:rewriter orchestrator (via /bookwright:revise) owns that flow. This agent drafts from plans.

Phase 1: Read the Plan

Read the chapter plan file in full before dispatching any subagent. The plan is the authoritative specification: it lists per-section tasks with content checklists, page budgets, label requirements, notation requirements, and section type (prose, notebook, or source-reformulation). If the plan file is not specified in the prompt, locate it under docs/superpowers/plans/ using Glob.

Also read:

  • book/CLAUDE.md for repo layout, naming conventions, build command, and style rules
  • docs/superpowers/bookwright.config.yaml for project-level settings (stack, macro package, etc.)
  • The two or three sections immediately preceding the target chapter for voice continuity

If the scope is ambiguous (chapter number, which plan file, single section vs. full chapter), use AskUserQuestion before proceeding.

Phase 2: Dispatch Drafting Specialists

For each section task in the plan, dispatch the appropriate specialist via Task:

Task type Specialist
Prose section bookwright:section-writer
Notebook / code companion bookwright:notebook-author
Content drawn from source papers bookwright:source-reformulator (then pass its output to section-writer)

Pass the full task spec from the plan, the relevant surrounding context, and any output from source-reformulator if applicable. Independent sections can launch in parallel. Sections that depend on a prior section's definitions must wait for that section's commit before launching.

Read the full file on GitHub · 106 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. 3d ago First seen · 106 lines · 368 tokens per session scan A bc8a46d344c8

Subscribe to this mod's changes

writer is an agent published in the GitHub repository queelius/claude-anvil (2 stars, last pushed 1mo ago), licensed MIT. It adds 368 tokens to every session and 1,407 once invoked, about $0.0018 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 2026-08-31.

Related

Other agents, from other repositories