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 xiamenhewei/zishen --skill zishengit clone --depth 1 https://github.com/xiamenhewei/zishenWrote 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/xiamenhewei/zishen/zishen)<a href="https://agentmods.dev/skills/xiamenhewei/zishen/zishen"><img src="https://agentmods.dev/badge/skills/xiamenhewei/zishen/zishen/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/xiamenhewei/zishen/zishen"><img src="https://agentmods.dev/badge/skills/xiamenhewei/zishen/zishen.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 | $0.00123 | $0.04614 |
| Opus 5 | $0.00062 | $0.02307 |
| Sonnet 5 | $0.00025 | $0.00923 |
| Haiku 4.5 | $0.00012 | $0.00461 |
Grade A, and why
自审 scanned grade A with 1 finding 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 8d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
每个改动点是否都有对应的验证方法?不是最后统一验一次,是每个改动同步想验证。**验证必须实跑**:浏览器打开看、curl 打一下、控制台看报错。禁止只看代码说"应该没问题"。 How it starts
The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.
自审框架
十四条审查 + 一条验收闭环,逐条给结论:通过 / 不通过 / 需补充。最后汇总。
铁规则
自审发现遗漏/不通过时,必须停下来等用户确认后再施工,禁止自审修正后直接开工。
自审通过也不能直接开工,必须等用户明确说"去做/开工/执行/改"才能动代码。自审通过 = 汇报结论等命令,不等于开工许可。
流程:自审 → 汇报结论 → 等用户说"去做" → 才能动代码。
清单
1. 需求符合性
方案是否完全覆盖用户提出的需求?有没有多做的、少做的、歪曲的?
2. 第一性原理
方案是否从根本问题出发?是否足够优雅?有没有过度设计?
3. 遗漏和错误
有没有遗漏的边界条件、异常场景、部署环境差异?
4. 客观事实依据
是否基于实际代码和数据确认?是否用代码索引工具/grep 看清了代码和依赖?是否验证过信源可靠性?禁止凭记忆猜测。
5. 复用已有模块
相同功能是否复用已有代码?是否造成同一功能多处维护?
参照物逐项对比(铁规则):当方案说"复用/参照/对齐/复现"某个已有模块或模式时,自动触发本规则,必须:
- 列出参照物的完整功能清单(每个功能点、每个细节)
- 逐项确认当前方案是否覆盖
- 找出差距,全部标注为"待补充"
- 禁止只看"表面差异"就认为对齐完毕——参照物里的每个功能都是必要项,除非有明确理由排除
反面教材:说"对齐某筛选面板",只看了"标题变化、面板关闭、高亮"三个表面差异,没有列出该面板的完整功能(滚动锁定、锚点定位、滚动位置补偿、URL 参数、浏览器前进后退恢复、跨窗口通知),导致漏了 6 个功能点。
跨域分别找参照物(铁规则):功能跨多个领域时(如"提交流程 + 图片处理 + 通知合并 + 手机号脱敏"),每个领域各自找该领域的标杆参照物,不能只对一套参照物验收就收——A 域的参照物清单里没有 B 域的功能项,照抄 A 就把 B 域"裸奔"也照抄了。
自审时问:方案涉及几个领域?每个领域的参照物分别是谁?有没有哪个领域拿了"根本不处理这个领域"的参照物(如图片处理去抄一个只传 PDF 的模块)?
反面教材:某报事提交入口"仿某函件提交模块",提交流程对齐没问题,但图片处理也照抄函件模块——函件传 PDF 根本不处理图片,于是没压缩、没水印、没大小限制,用户拍照三四兆(MB,文件大小单位)裸存。图片处理的真正参照物是项目里的图片上传模块(有压缩 + 质量参数 + 水印 + 大小上限全套)。自审只对照了"提交流程"这套参照物,漏了"图片处理"那套。
6. 长期最优解
是否考虑数据规模增长?是否选了长期可维护的方案而非临时凑合?
7. 根本问题
是解决了根本原因,还是绕过表面症状?
8. 验证方法
每个改动点是否都有对应的验证方法?不是最后统一验一次,是每个改动同步想验证。验证必须实跑:浏览器打开看、curl 打一下、控制台看报错。禁止只看代码说"应该没问题"。
9. 影响范围
改动的代码被谁引用/依赖?是否用代码索引/grep 追踪了所有引用点确保全覆盖?
DOM 操作检查:涉及 getElementById / querySelector + addEventListener 的代码,必须确认 DOM 插入(appendChild/innerHTML)在事件绑定之前。否则 getElementById 返回 null 导致整段 JS 崩。
显示类漏点搜索(铁规则):找"显示某字段(时间/手机号/哈希等需脱敏或统一格式字段)"的漏点时,不能只搜格式化函数名(如 formatTime / maskPhone)——有的地方用 innerHTML / innerText 直接吐原始字段值,根本不套任何格式化/脱敏函数,搜函数名永远搜不到它们。必须搜"字段名 + 输出动作"组合(如先 grep 字段名,再看附近有没有 innerHTML 直接吐),把所有"显示这个字段"的点列全再逐个确认。反面教材:要给某时间字段统一脱敏,只搜格式化函数找漏点,漏了两处 innerHTML(row.created_at) 直接吐原始时间——这俩没套脱敏函数,搜函数名是盲区。
数据字段语义改动(铁规则):改数据写入/字段语义(如时间字段从 created_at 改成 in_time、状态码改义)时,影响范围是数据流视角不是调用链视角——代码调用关系只覆盖"谁调了这个函数",覆盖不了"并行的查询接口自己 SELECT 了这个字段"。必须 grep -rn "字段名" *.py 列出全仓库所有 SELECT/读取该字段的地方,逐个确认要不要同步改。区分:改函数逻辑查调用链;改数据字段语义查 grep SELECT(数据消费链,易漏)。反面教材:改某按日聚合的日期口径,只查写入函数的调用方,漏了一个对外查询接口自己 SELECT 原始事件表按 created_at 显示,结果改了明细卡片没读、数据还堆在旧日期。
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.
- 8d ago Changed · +46 lines 839ad3106ff7
- 11d ago First seen · 104 lines · 123 tokens per session scan A 913006f3b22c
自审 is a skill published in the GitHub repository xiamenhewei/zishen (2 stars, last pushed 11d ago), licensed MIT. It adds 123 tokens to every session and 4,614 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
omh-code-review
This is a Hermes-native code-review workflow skill.
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
full-repo-review
Comprehensive four-wave review of all repo source files, producing a prioritized issue backlog.
agent-teams-simplify-and-harden
Implementation + audit loop using parallel agent teams with structured simplify, harden, and document passes. Spawns implementation agents to do the work, then audit agents to find complexity, security gaps, and spec deviations, then loops until code compiles cleanly, all tests pass, and auditors find zero issues or…