Marin is an open-source research program, software platform, and community for developing foundation models such as large language models. Researchers use it for data preparation, tokenization, pretraining, posttraining, evaluation, and related experiments, including work on audio-text, DNA, and protein models. The catalogue entries are add-ons that support workflows around Marin.
Borrowing it
Nothing to install: this file belongs to marin-community/marin. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/marin-community/marin/main/.agents/skills/writing-style/SKILL.mdgit clone --depth 1 https://github.com/marin-community/marinWrote 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/marin-community/marin/writing-style)<a href="https://agentmods.dev/skills/marin-community/marin/writing-style"><img src="https://agentmods.dev/badge/skills/marin-community/marin/writing-style/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/marin-community/marin/writing-style"><img src="https://agentmods.dev/badge/skills/marin-community/marin/writing-style.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00029 | $0.00468 |
| Opus 5 | $0.00015 | $0.00234 |
| Sonnet 5 | $0.00006 | $0.00094 |
| Haiku 4.5 | $0.00003 | $0.00047 |
Grade A, and why
writing-style 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 10d 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 — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Marin House Style
Read the medium-specific guide for the deliverable, then apply ai-writing-donts.md as the final prose pass.
Read The Right File
- Read blog-posts.md for public explainers and blog posts.
- Read reports.md for technical reports aimed at peer researchers outside Marin.
- Read tutorials.md for learning-oriented documentation that introduces Marin.
- Read reference-docs.md for precise usage docs aimed at readers who already know Marin.
- Read issues.md for standard OSS issues and experiment issues.
- Read pull-requests.md for commit messages and PR titles and bodies.
- Read discord.md for Discord summaries and tactical replies.
- If a piece spans media, use the stricter applicable rule.
Voice
Write as a rigorous open-science lab building foundation models. Treat experiments, decisions, and mistakes as part of the record. Use a sober, practical voice for technically capable peers. Be natural and direct; use specialized language when it improves precision.
On Voice
These rules are stricter for agents. Preserve harmless human flourishes when editing another person's prose.
Evidence and scope
- Lead with the result, decision, or changed behavior.
- Support claims with numbers, examples, observations, and relevant artifacts.
- State uncertainty and limitations plainly; do not imply broader evidence.
- Remove hype, product-announcement language, AGI speculation, decorative emoji, and adjectives standing in for evidence.
- Explain non-standard terms. Add detail only when it helps the reader reproduce, evaluate, or act.
Add Structure For Longer-Form Writing
Long-form pieces may need a document-level TL;DR and section takeaways. Do not force them into short media.
Before publishing, confirm the prose is factually current, sounds normal when spoken to a colleague, cites material evidence, and contains no generic filler.
What ships with it
8 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.
- 10d ago First seen · 51 lines · 29 tokens per session scan A c03fc281b0aa
writing-style is a skill published in the GitHub repository marin-community/marin (3,548 stars, last pushed today), licensed Apache-2.0. It adds 29 tokens to every session and 468 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.
Other skills, from other repositories
create-pr
Creates a GitHub PR with a Linear-ticket-prefixed title and a decision-led, narrative description for prisma-next. Use when the user wants to create a pull request, open a PR, or submit changes for review.
git-commit
Generate well-formatted git commit messages following conventional commit standards.
repo-sync
Manage repo-sync on macOS for shared documents and team context. Use when asked what repo-sync does, to automatically sync a repo, add a repo to repo-sync, stop syncing or un-sync a repo, check sync health, configure repo-sync, uninstall it, or install and maintain its bundled agent skill. Ordinary one-time Git pulls…
mcore-split-pr
Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.
contributing
How to contribute to evlog, covering commit and PR conventions, changesets, the Definition of Done, testing rules, and the authored skills that walk through building a new adapter, enricher, framework integration, or map rule. Load this for any question about contributing, opening a PR, or adding something to the…
strict-tdd
Strict RED->GREEN->REFACTOR test-driven development with enforcement. Never write production code before a failing test. Atomic commits per TDD cycle.