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.
npx skills add sidiangongyuan/codex-skills-library --skill paper-writinggit clone --depth 1 https://github.com/sidiangongyuan/codex-skills-libraryWrote 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.
[](https://agentmods.dev/skills/sidiangongyuan/codex-skills-library/paper-writing)<a href="https://agentmods.dev/skills/sidiangongyuan/codex-skills-library/paper-writing"><img src="https://agentmods.dev/badge/skills/sidiangongyuan/codex-skills-library/paper-writing/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sidiangongyuan/codex-skills-library/paper-writing"><img src="https://agentmods.dev/badge/skills/sidiangongyuan/codex-skills-library/paper-writing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00041 | $0.01448 |
| Opus 5.5 | $0.00016 | $0.00579 |
| Sonnet 5.5 | $0.00008 | $0.00290 |
| Haiku 4.5 | $0.00004 | $0.00145 |
Grade A, and why
paper-writing 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 26d 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.
How it starts
The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Paper Writing
Make the paper's purpose, contribution, use, and evidence easy to follow. Adapt the writing to the actual paper rather than fitting it to a method-paper template.
Choose the scope before the workflow
- Section writing: draft or restructure the requested sections. Read their relevant guides below and identify the paper's main contribution before editing.
- Local refinement: read the target and its immediate context, then use prose.md. Preserve the settled story and verify affected links only. A caption correction does not require a new literature survey, contribution interview, full manuscript rewrite, or experiment run.
- Whole-manuscript consistency: read consistency.md and trace claims and terminology across the main text, visuals, and supplement. Use a section guide only where a concrete structural issue needs repair.
An assessment request authorizes diagnosis, not file changes. An explicit revision request authorizes scoped edits and their normal verification. Missing experiments are evidence gaps, not permission to launch jobs. Ask only when a decision changes the scientific claim, scope, resources, or an external action; inspect existing records first. Do not reopen settled choices without a concrete contradiction.
Establish the argument
For structural work, identify from the draft and evidence:
- the reader's question and the paper's primary contribution;
- what the resource, method, or analysis enables, and how it is actually used;
- which evidence supports each principal claim, and what remains untested;
- the role of reference implementations, auxiliary supervision, and controls.
Keep this working map compact; it need not become a new document. For a benchmark paper, lead with the resource and what its use demonstrates, not a compulsory model invention. For a method paper, connect the problem to the tested design. For a system paper, connect requirements and interfaces to end-to-end evidence. For an analysis paper, connect the question and measurement to the finding. Hybrid papers may combine roles without forcing every component into an equal contribution. Do not invent a new name or mechanism to create a narrative turn.
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 26d ago Changed · +29 lines 2688d1786ea1
- 1mo ago First seen · 96 lines · 41 tokens per session scan A 8cae9b114b74
paper-writing is a skill published in the GitHub repository sidiangongyuan/codex-skills-library (25 stars, last pushed 26d ago), licensed MIT. It adds 41 tokens to every session and 1,448 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-15.
Other skills, from other repositories
zeroplot
Create editable charts, diagrams, figures, posters, or slides from intent, data, or references.
nature-reviewer
A tool for simulating peer review of a research paper from the reviewer’s point of view. Peer review is the process in which experts assess a paper before publication.
nature-ref-verifier
A reference-checking workflow that compares bibliographic details across multiple sources. Bibliographic details include authors, title, year, volume, issue, pages, and DOI, the persistent identifier for a research publication.
nature-statistics
A guide for checking or writing how statistical analyses are reported in scientific papers submitted to Nature or similar high-impact journals.
nature-paper-trans
A workflow for translating selected pages or a complete English research-paper PDF into a simplified-Chinese, image-based PDF. bioRxiv is not mentioned in the provided details; the workflow applies to research-paper PDFs generally.
nature-experiment-log
A workflow for turning experiment notes, images, audio, or text into structured Markdown laboratory logs with YAML metadata. It can also organize raw attachments and optionally connect the logs to Feishu or Obsidian.