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 agentmods add skills/zhinjs/zhin/plugin-qualitynpx skills add zhinjs/zhin --skill plugin-qualitygit clone --depth 1 https://github.com/zhinjs/zhinWhat 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 | $0.00049 | $0.00931 |
| Opus 5 | $0.00024 | $0.00465 |
| Sonnet 5 | $0.00010 | $0.00186 |
| Haiku 4.5 | $0.00005 | $0.00093 |
Grade A, and why
plugin-quality scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Zhin 插件质量审查(Plugin Runtime)
对照 Plugin Runtime 审查插件:结构、约定目录、依赖方向、发送链与安全。发现仍用 usePlugin / MessageCommand 的新代码应标为缺陷并迁移。
适用场景
- 「检查插件质量」「审查代码」「发布前看看有没有问题」
- 重构现有插件使其符合 Runtime
审查维度
1. 结构规范性
| 项目 | 要求 | 严重程度 |
|---|---|---|
| 入口 | plugin.ts default-export definePlugin() |
🔴 严重 |
| manifest | package.json#zhin(protocol、entry、features) |
🔴 严重 |
| 能力放置 | 约定目录 + default export;非业务堆在 plugin.ts |
🔴 严重 |
| 旧 API | 新代码无 usePlugin / getPlugin / MessageCommand |
🔴 严重 |
| 导入路径 | 相对导入带 .js |
🔴 严重 |
| Feature 依赖 | 用到的能力已在 zhin.features 声明 |
🟡 中等 |
| 配置 | 自有字段在 schema.json,经 context.config 读取 |
🟡 中等 |
常见错误:
// ❌ 经典插件 API(Runtime 下不工作)
import { usePlugin, MessageCommand } from 'zhin.js'
const plugin = usePlugin()
plugin.addCommand(new MessageCommand('hi').action(() => 'hi'))
// ✅ Runtime
import { definePlugin } from 'zhin.js'
export default definePlugin({ name: 'my-plugin', setup() {} })
// + commands/hi.ts → defineCommand(...)
2. 生命周期与资源
- 清理走
context.lifecycle.add或setup返回 disposer - 共享连接/DB 用
context.resources.provide,禁止裸模块级单例或createGenerationStore - Host token 先
has再use(精简安装下可选)
3. 发送与安全
- 出站不绕过
Message.$reply/Adapter.sendMessage→before.sendMessage - 密钥只走环境变量 /
${VAR}配置引用 - AI 工具副作用边界清晰;危险操作有策略/确认
4. 测试与文档
- 至少覆盖核心
execute路径 - README 写明命令触发方式与 Feature 依赖
- 带
agent/的包检查files/prepublishOnly(pnpm check:plugin-agent-publish)
审查流程
- 读
package.json#zhin+plugin.ts - 扫
commands//tools//middlewares/是否 default export 正确 API - ripgrep:
usePlugin|MessageCommand|getPlugin|bootstrapNode - 跑
pnpm --filter <pkg> test(及需要的 harness) - 输出问题清单(严重度 + 建议修复)
输出格式
## 结论
- 是否可发布:是 / 否
## 问题
| 严重度 | 位置 | 问题 | 建议 |
|--------|------|------|------|
| 🔴 | ... | ... | ... |
## 已通过
- definePlugin 入口 / 约定目录 / 发送链 …
迁移
旧插件迁 Runtime:.github/skills/migrate-zhin-plugin-runtime。
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.
- 2d ago First seen · 100 lines · 49 tokens per session scan A 84f018dddb7a
plugin-quality is a skill published in the GitHub repository zhinjs/zhin (135 stars, last pushed 5d ago), licensed MIT. It adds 49 tokens to every session and 931 once invoked, about $0.0002 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
test-driven-development
TDD: enforce RED-GREEN-REFACTOR, tests before code.
military-discipline
严格执行,质量把关,拒绝敷衍.
code-review-skill
Review code for quality issues, bugs, and adherence to best practices.
code-review
Comprehensive code review checklist for pull requests.
code-review
Code review for issues, style consistency, best practices—use for full code evaluation; not for debugging specific errors or fixing individual style violations.
code-review
Comprehensive code review for security, performance, and best practices.