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 orziz/odai --skill renweigit clone --depth 1 https://github.com/orziz/odaiWrote 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/orziz/odai/renwei)<a href="https://agentmods.dev/skills/orziz/odai/renwei"><img src="https://agentmods.dev/badge/skills/orziz/odai/renwei/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/orziz/odai/renwei"><img src="https://agentmods.dev/badge/skills/orziz/odai/renwei.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 | $0.00109 | $0.00954 |
| Opus 5.5 | $0.00044 | $0.00382 |
| Sonnet 5.5 | $0.00022 | $0.00191 |
| Haiku 4.5 | $0.00011 | $0.00095 |
Grade A, and why
renwei 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.
What it actually says
Renwei
“人味”：把一眼能看出批量生成的内容，改成有主人、有受众、有分寸的作品。它负责表现工艺，不替用户决定目标、受众和品牌取舍，也不帮人隐瞒必须披露的 AI 使用。
先看清味从哪来
- 从实际交付物出发：上线的文案、商店页、包体里的图、成片和缩略图，不只看源稿或提示词。先盘点哪些东西真的会被人看到。
- 逐项指出具体症状和出现位置，比如套话、模板构图、假字、噪点、风格割裂；按“影响大、改动小”排序。不用“整体感觉”代替证据。
- 找到真样本：产品自己的画面、已认可的文案、真实截图与录像、品牌规范。新内容向真样本对齐，不向生成器的默认风格对齐。
原则
- 具体胜过华丽：用真实的名词、动作、数字和画面，删掉撑不起事实的形容和口号。
- 一致胜过精致：一组素材、一套文案要像同一只手做的。宁可朴素统一，也不要单件惊艳、整体拼贴。
- 真实胜过摆拍：宣传优先用产品本身（实机、真实截图、原有音乐与美术）来组合，不另造模板化场面。
- 少而好：同质内容越多越稀释，能删就删，能合并就合并。
- 不属于你的原文不动：合作方提供的、授权的、法务的、引用的文本和素材逐字保留，润色范围之外不碰。
- 不歪曲：润色不改事实、数值、承诺与范围；本地化译意，不加戏。
做法
- 确认受众、用途、载体、尺寸或字数限制、交付格式，以及保存、上传、发布是否在授权内。
- 诊断，定真样本，改写或重做，再在最终形态里验收：看实际页面、实际尺寸、实际包体或成片。具体工艺按需读取：
- 文案、口播、界面文字、本地化：
references/copy.md - 图像、界面素材、宣传图、截图、视频、噪点、生成来源标记：
references/visual.md
- 文案、口播、界面文字、本地化：
- 借其他模型或工具干活时给出明确约束，并逐条审查回交结果。新添的口号、套话、额外含义和风格漂移一律退回。
- 交付时说明改了什么、依据哪些真样本、怎么验收的，以及哪些没法验证。
边界
- 目的是质量和诚实，不是骗过 AI 检测：不宣称“零 AI”“纯人工原创”，不删除平台或法规要求的 AI 使用披露和必需的版权信息。
- 生成来源元数据（如 C2PA / Content Credentials）只在交付副本上清理，原件和来源记录保留。像素或信号里的内嵌水印（如 SynthID）不能靠转码、截图、裁切证明已去除；需要时换用无水印来源，不宣称已去除。
- 对外发布、上传、替换线上内容仍以用户授权为准；本地文件改好不等于线上已更新。
若由 odai 调用，返回改动、依据、验收证据与未决项，由 odai 统一核对目标、边界与最终交付。
What ships with it
3 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 First seen · 41 lines · 109 tokens per session scan A 48b3118cb414
renwei is a skill published in the GitHub repository orziz/odai (119 stars, last pushed yesterday), licensed MIT. It adds 109 tokens to every session and 954 once invoked, about $0.0004 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 2026-10-26.
Other skills, from other repositories
dsh-run-output
Internal guidance for presenting DeepSeek Harness bridge results (reviews, critiques, delegated runs) back to the user. Read before summarizing any dsh-bridge.mjs output.
dev-qa
A development and quality process for building software, testing it, and checking it for security problems. It separates implementation, quality assurance, and security review while coordinating their handoffs.
promotion
A release-stage workflow for promotion and early user adoption. It asks the user what promotional materials to make before creating anything, and its instructions are written in Chinese.
idea-validation
A structured process for checking whether a product idea is worth building. It researches the market, competitors, and technical feasibility, then records a recommendation and the assumptions behind it.
product-prd
A Chinese-language product-definition and requirements workflow that turns an approved idea into a product definition and a PRD, a detailed product requirements document.
design
A technical design stage that turns a product brief into an implementation plan and a list of small, checkable development tasks.