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 travsteward/openwriter --skill skillgit clone --depth 1 https://github.com/travsteward/openwriterWrote 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/travsteward/openwriter/skill)<a href="https://agentmods.dev/skills/travsteward/openwriter/skill"><img src="https://agentmods.dev/badge/skills/travsteward/openwriter/skill/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/travsteward/openwriter/skill"><img src="https://agentmods.dev/badge/skills/travsteward/openwriter/skill.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 | $0.00180 | $0.17994 |
| Opus 5.5 | $0.00072 | $0.07198 |
| Sonnet 5.5 | $0.00036 | $0.03599 |
| Haiku 4.5 | $0.00018 | $0.01799 |
Grade A, and why
openwriter 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 today.
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.
Copies of this mod
1 near-identical copy found in the catalogue:
- openwriter — 94% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 817 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OpenWriter Skill
Active harness
This skill is shared by every agent harness. Harness-specific execution (how to run a background worker, which browser tool to use) lives in one short document per harness; read only the one for the harness you are running in:
- Claude Code: docs/harness-claude.md
- Codex: docs/harness-codex.md
- Setup for Claude Code, Codex and OpenCode: docs/setup.md
Tool names below are logical names: discover the current MCP namespace and input
schema. Never copy a Claude Agent(...) call from a server footer into another
harness.
Revision variants
Use Create variant → Revision on the source document. The revision nests under its parent's chevron and retains its writing format. A manuscript revision compiles accepted source text into one ordinary editable document. Unresolved references stop creation. The source and its pending suggestions stay intact. Later source changes do not flow into the revision. The copy starts with normal pending review and an original-copy version to restore.
For agents, create_editing_draft({ docId, title? }) is the compatibility name
for this same operation. It returns identity and chapter headings, never the
book body. Do not follow it with populate_document. Creation is background;
scope subsequent reads and writes to the returned docId.
Read through outline_doc, then peek_doc or read_pad slices for the chapter
being edited; don't force-read the entire book for routine edits. Read and mark
up the revision in the normal editor's Focus mode, where selections, comments,
and edits remain available. No separate reading copy or reading surface is needed.
You are a writing collaborator. You read documents and make edits exclusively via MCP tools. Edits appear as pending decorations (colored highlights) in the user's browser that they accept or reject.
FIRM RULES
- NEVER OPEN THE OPENWRITER UI (localhost:5050) IN A BROWSER — Chrome, the browser pane, or any other. The browser is the user's view; every tab shares the server's one open document, so an agent tab opening a doc moves the user's screen mid-review. Read and verify through MCP tools only (
read_pad,peek_doc,get_pad_status). Need the user to look at something? Tell them the docId, orswitch_documenton their explicit request. Checking a published page (the live blog URL) in a browser is fine.
What ships with it
2 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.
- today Changed 492047a691da
- yesterday Changed c7219e01f12b
- 11d ago Changed · +1 tokens per session 413314e7ef31
- 12d ago Changed · +1 lines 93279ba3e488
- 13d ago Changed · +57 lines 518971d78a97
- 14d ago Changed · -4 lines 5d0a9a61033b
- 19d ago Changed · +62 lines 9bdce665d9c8
- 21d ago Changed · +5 lines · -5 tokens per session cbbe9ac68f7a
- 1mo ago Changed · +17 lines 56cfc7036eca
- 1mo ago First seen · 679 lines · 184 tokens per session scan A 0c6c7a6f4e69
openwriter is a skill published in the GitHub repository travsteward/openwriter (36 stars, last pushed today), licensed MIT. It adds 180 tokens to every session and 17,994 once invoked, about $0.0007 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 2026-08-30.
Other skills, from other repositories
humanize-pro
Use when the user asks to humanize, unslop, de-slop, de-AI, clean up, tighten, or make text sound human or natural; when they say something reads like AI or like ChatGPT; when they ask whether something matches their voice or style guide; or when drafting or rewriting outbound human communication: LinkedIn posts and…
anti-ai-writing
Use when writing or editing any prose for a human reader - documentation, READMEs, emails, reports, captions, scripts, blog posts, UI copy, commit messages, PR descriptions - or when asked to remove AI tells, de-slop text, make writing sound human, or check writing style. Applies 29 researched patterns that make text…
humanizer-workbench
Rewrites AI-generated text into natural, human-quality writing using a multi-stage detection and transformation pipeline.
anti-slop
Remove AI slop from any voice-bearing prose — original posts, threads, articles, long-form, emails, docs, READMEs, marketing copy, bios, scripts. Use when drafting text meant to sound like a specific person or brand, when rewriting text that reads generic, corporate, or AI-generated, or when asked to humanize…
writing-humanizer
A text-editing skill that removes patterns often associated with AI-written prose and makes writing sound more natural. Its instructions are aimed mainly at Traditional Chinese text used in Taiwan.
kill-slop
Audit a file or draft for AI slop and off-voice lines against the author's bound voice, report findings with in-voice swaps, then apply approved fixes. Use when the user says "kill slop", "/kill-slop [file]", "find the slop in this", "audit this for slop", or wants a deck, article, page, or draft cleaned to sound like…