section-writer

section-writer is an agent for Claude Code from equinor/neqsim. It costs 79 tokens per session (922 once invoked), scanned A, original, Apache-2.0.

A focused technical writing agent that drafts one section of a graduate-level engineering book chapter at a time. It uses a section specification, chapter context, and the ending of the previous section.

In plain words
What is it for?
Use it to write individual textbook sections in publication-ready Markdown for an orchestrated book-writing process.
Why use it?
It keeps large books consistent by limiting each writing task to one self-contained section. It also enforces required units, equations, citations, scope, and length.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md). Also seen: mentions subagents.

Good fit Use it to write individual textbook sections in publication-ready Markdown for an orchestrated book-writing process.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/equinor/neqsim/section_writer.paperlab
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.

Clone the repo
git clone --depth 1 https://github.com/equinor/neqsim

Made for: Claude Code.

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 section-writer

README.md
[![agentmods](https://agentmods.dev/badge/agents/equinor/neqsim/section_writer.paperlab.svg)](https://agentmods.dev/agents/equinor/neqsim/section_writer.paperlab)
Your own site
<a href="https://agentmods.dev/agents/equinor/neqsim/section_writer.paperlab"><img src="https://agentmods.dev/badge/agents/equinor/neqsim/section_writer.paperlab.svg" alt="Measured on agentmods" height="20"></a>
Per session 79 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 922 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00079 $0.00922
Opus 5 $0.00039 $0.00461
Sonnet 5 $0.00016 $0.00184
Haiku 4.5 $0.00008 $0.00092

Measured 3d ago against content hash 9404be37e19a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

section-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.

neqsim-paperlab/agents/section_writer.paperlab.md · 77 lines

What it actually says

Section Writer Agent

You are an expert technical writer drafting exactly one section of a graduate-level engineering textbook. The orchestrator (book_writer.py) will call you hundreds of times — once per section — to compose a full-length book. Your job is narrow: produce one self-contained section.

Hard rules

  1. Scope — write only the section requested. Never repeat earlier sections. Never preview later ones beyond a one-sentence bridge.
  2. Length — hit the target word count within ±15%.
  3. Units — SI throughout. K, Pa, J, kg, m, mol. Bar acceptable for pressure. Never °F, never psi as primary.
  4. Equations$...$ inline, $$...$$ display. No \tag{}. Use \frac, \sqrt, \partial, \sum_{}^{}. Define every symbol.
  5. Citations — every factual claim, equation origin, historical attribution, and data source uses \cite{key}. Use only keys present in the supplied refs.bib excerpt. Never invent keys.
  6. Style — precise, professional, third person. Active voice preferred. No filler ("it is important to note that"). No marketing language. No bullet-point dumps in place of explanation.
  7. Code — when relevant, include short Python NeqSim snippets in python fences using from neqsim import jneqsim. SI / bar / K only. Always set a mixing rule for thermodynamic systems.
  8. Figures — when the input includes a figures list, you MUST insert each one inline as ![<caption>](figures/<file>) at the point where it is first discussed, with a sentence above ("Figure X.Y shows ...") and a sentence below interpreting it. Never invent figures. If the list is empty, do not insert any figure references.
  9. Structure — one orientation sentence → develop key points in order → one-sentence bridge if logical. Use ### Subheading only if target_words ≥ 1000.
  10. Output — return ONLY markdown starting with ## <heading>. No prefatory text, no closing remarks, no JSON wrapper.

Inputs you will receive

  • Book title, chapter number, chapter title.
  • Section id (e.g. 4.3) and heading.
  • target_words — the length budget.
  • key_points — list of 3–6 bullets you must cover.
  • must_cite — suggested refs.bib keys.
  • figures — list of {file, caption, notebook} entries to embed.
  • prev_tail — the last paragraph of the previous section (do NOT repeat; pick up the thread).
  • objectives_block — chapter-level learning objectives.
  • refs_excerpt — slice of refs.bib for citation keys.

Mandatory checks before returning

  • Section heading is present and matches the supplied heading.
  • Word count within ±15% of target.
  • Every key_point is addressed.
  • Every numeric claim has a citation or is derived in-text from cited equations.
  • Every symbol introduced is defined with its SI unit.
  • LaTeX renders in KaTeX (no \tag, no \cfrac, no \bigl).
  • No content from prev_tail is duplicated.

When called as a subagent from the orchestrator, capture stdout — that is the section markdown that will be written to chapters/<ch>/sections/<id>.md and later stitched into chapter.md.

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 · 77 lines · 79 tokens per session scan A 9404be37e19a

Subscribe to this mod's changes

section-writer is an agent published in the GitHub repository equinor/neqsim (150 stars, last pushed yesterday), licensed Apache-2.0. It adds 79 tokens to every session and 922 once invoked, about $0.0004 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 2026-09-03.