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 LiPu-jpg/Openwrite --skill dog-delivery-verificationgit clone --depth 1 https://github.com/LiPu-jpg/OpenwriteWrote 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/lipu-jpg/openwrite/dog-delivery-verification)<a href="https://agentmods.dev/skills/lipu-jpg/openwrite/dog-delivery-verification"><img src="https://agentmods.dev/badge/skills/lipu-jpg/openwrite/dog-delivery-verification/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/lipu-jpg/openwrite/dog-delivery-verification"><img src="https://agentmods.dev/badge/skills/lipu-jpg/openwrite/dog-delivery-verification.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 | $0.00081 | $0.01822 |
| Opus 5.5 | $0.00032 | $0.00729 |
| Sonnet 5.5 | $0.00016 | $0.00364 |
| Haiku 4.5 | $0.00008 | $0.00182 |
Grade A, and why
dog-delivery-verification 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.
How it starts
The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
DoG 独立验收(OpenWrite → dsh-dog)
这条技能把 OpenWrite 的正文与评审证据交给宿主已安装的 @dsh-external/dsh-dog,
由它用独立 verifier 再判断一次。它不是 OpenWrite 六域评审的替代品,也不修改正文。
先确认工具面真的有 dog_validate / dog_create / dog_run。没有时不要假装跑过:
说明 dsh-dog 未安装或当前宿主不放行,回到 novel_review_chapter 与
novel_export(action="preflight") 的 OpenWrite 门禁。
1. 收集被测对象与证据
单章:
novel_outline_read与novel_status定位chapter_id。novel_chapter_work读取正文 revision、评审新鲜度、本章义务。novel_review_chapter(action="versions", chapter_id=...)取最新评审; 若正文 revision 与评审 revision 不一致,先重新评审,不要把旧评审当证据。novel_export(action="preflight", purpose="delivery")读取交付门禁和缺口。
整书:先 novel_status,再对每个已写章节重复上面的 2–4;不要把未写章节算进范围。
2. 建立不可变交付包
在当前工作区根目录下创建 .openwrite/dog-handoff/<handoff-id>/(示例用 1),写入:
manuscript.md:被测正文的逐字副本(单章就是该章;整书按章节分隔)。evidence.md:outline 摘要、最新评审的 revision/分数/门禁/问题、交付 preflight 缺口。graph.json:下面的 0.9 图。target必须是相对会话工作区根目录的路径,不是相对 handoff 目录; root 指向 handoff 目录(dog 会把目录打包捕获),叶子分别指向其中的manuscript.md/evidence.md。
用文件工具写,不要用 novel_doc 写项目外路径。handoff-id 用时间戳或短随机串,
保留目录供审计;不要把它放进 data/manuscript/。
3. 图模板
{
"schemaVersion": "0.9",
"id": "openwrite-delivery-1",
"root": "root",
"nodes": {
"root": {
"kind": "composite",
"title": "OpenWrite 交付整体判断",
"constraint": "hard",
"target": ".openwrite/dog-handoff/1/",
"completion": { "op": "all", "items": [
{ "op": "ref", "id": "continuity" },
{ "op": "ref", "id": "prose" },
{ "op": "ref", "id": "canon" },
{ "op": "ref", "id": "delivery-evidence" }
] },
"verifier": {
"mode": "agentic",
"instruction": "Read manuscript.md and evidence.md in the captured handoff directory. Judge whether the chapter(s) hold together as a deliverable: continuity, prose quality, canon consistency and the claimed delivery evidence. Return pass only when the evidence supports every claim; otherwise fail with the exact gap."
}
},
"continuity": {
"kind": "leaf", "title": "连续性", "constraint": "hard", "target": ".openwrite/dog-handoff/1/manuscript.md",
"verifier": { "mode": "agentic", "instruction": "Check scene-to-scene continuity in manuscript.md: who is present, where, when, what changed, and whether any action contradicts the previous scene. Cite exact lines." }
},
"prose": {
"kind": "leaf", "title": "文笔", "constraint": "hard", "target": ".openwrite/dog-handoff/1/manuscript.md",
"verifier": { "mode": "agentic", "instruction": "Judge prose quality in manuscript.md: viewpoint discipline, concrete detail, dialogue that carries character, and absence of filler or repeated exposition. Cite exact lines." }
},
"canon": {
"kind": "leaf", "title": "正典一致性", "constraint": "hard", "target": ".openwrite/dog-handoff/1/manuscript.md",
"verifier": { "mode": "agentic", "instruction": "Check manuscript.md against the facts in evidence.md: names, world rules, timeline and stated canon. Report any contradiction with the exact conflicting lines." }
},
"delivery-evidence": {
"kind": "leaf", "title": "交付证据", "constraint": "hard", "target": ".openwrite/dog-handoff/1/evidence.md",
"verifier": { "mode": "agentic", "instruction": "Check whether evidence.md actually supports delivery: review revision matches the manuscript, gate status is pass, and no listed preflight gap remains. Do not accept a claim that is not backed by the evidence text." }
}
},
"contains": [
{ "parent": "root", "child": "continuity", "required": true, "failure": "fatal" },
{ "parent": "root", "child": "prose", "required": true, "failure": "fatal" },
{ "parent": "root", "child": "canon", "required": true, "failure": "fatal" },
{ "parent": "root", "child": "delivery-evidence", "required": true, "failure": "fatal" }
],
"dependsOn": []
}
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 First seen · 114 lines · 81 tokens per session scan A 0b099dee4ece
dog-delivery-verification is a skill published in the GitHub repository LiPu-jpg/Openwrite (930 stars, last pushed today), licensed Apache-2.0. It adds 81 tokens to every session and 1,822 once invoked, about $0.0003 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 2026-11-07.
dog-delivery-verification compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| dog-delivery-verificationLiPu-jpg/Openwrite | Skill | 930 | today | Apache-2.0 | A | 81 | Claude Code |
| webnovel-initlingfengQAQ/webnovel-writer | Skill | 7.9k | today | GPL-3.0 | A | 41 | Claude Code |
| webnovel-writelingfengQAQ/webnovel-writer | Skill | 7.9k | today | GPL-3.0 | A | 30 | Claude Code |
| webnovel-planlingfengQAQ/webnovel-writer | Skill | 7.9k | today | GPL-3.0 | A | 33 | Claude Code |
| webnovel-reviewlingfengQAQ/webnovel-writer | Skill | 7.9k | today | GPL-3.0 | A | 24 | Claude Code |
| webnovel-querylingfengQAQ/webnovel-writer | Skill | 7.9k | today | GPL-3.0 | A | 35 | Claude Code |
| webnovel-learnlingfengQAQ/webnovel-writer | Skill | 7.9k | today | GPL-3.0 | A | 22 | Claude Code |