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 hanzhcn/laohan-skills --skill laohan-jiaozhungit clone --depth 1 https://github.com/hanzhcn/laohan-skillsWrote 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/hanzhcn/laohan-skills/laohan-jiaozhun)<a href="https://agentmods.dev/skills/hanzhcn/laohan-skills/laohan-jiaozhun"><img src="https://agentmods.dev/badge/skills/hanzhcn/laohan-skills/laohan-jiaozhun/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/hanzhcn/laohan-skills/laohan-jiaozhun"><img src="https://agentmods.dev/badge/skills/hanzhcn/laohan-skills/laohan-jiaozhun.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 | $0.00099 | $0.02038 |
| Opus 5.5 | $0.00040 | $0.00815 |
| Sonnet 5.5 | $0.00020 | $0.00408 |
| Haiku 4.5 | $0.00010 | $0.00204 |
Grade A, and why
laohan-jiaozhun 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 16d 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.
How it starts
The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.
内容校准适配器
合同地位（2026-09-20治理裁定）
本skill是校准适配器辅助工具，未登记executor-registry——④节点合同入口是上游cheat-on-content（registry node 4），⑤深扫合同入口是dbs-content-review+laohan-shencha-content-claims（node 5）。宿主按registry/whois执行④⑤，不因本skill目录可见而路由到它；本skill的lane路由、防污染与登记流程作为操作手册参考。
这里不是第二套 Cheat。cheat-on-content 是唯一校准引擎；本 skill 只把 episode 和它的校准 lane 接起来，避免本地 fork 公式、预测规则或迁移规则漂移。
不可跨越的边界
- 禁止在这里按固定教程型权重打分，禁止估算后把分数当预测，禁止另写 PREDICT / RETRO / BUMP。
- 禁止把
tutorial-video的历史样本拿来评估opinion-video；两种内容必须是独立 lane。 - 盲预测必须在拍摄、发布、查看本期实际数据之前，由上游
cheat-on-content写入其predictions/；预测段不可改写。 - 上游版本更新后若 state schema 不匹配，只能先执行
cheat-migratedry-run。备份与用户确认前不得迁移历史 lane。
路由
评分前的 verify-blind-rubric.mjs 不得只核对 header：blind-rubric.md 必须逐字保留来源 rubric 当前完整维度定义与占位规则。维度语义有缩写、遗漏或改写时保持 PENDING，修复后换全新隔离上下文重评。
- 先运行
laohan-bianpai vendors。Cheat 或 dbskill upstream 有更新时只记录待办；只在新 episode 前或独立维护窗口运行vendors --sync。它报告 schema migration 时停止，不能绕过，也不能在本期中途切换规则。 - 读取
episodes/<slug>/01-口播稿.md与00-选题.md，判断内容形态。- 教程、明确操作路径、资源交付 →
tutorial-videolane。 - 观点、评论、议题、个人判断 →
opinion-videolane。 - 无法判断时停止，不能默认为教程型。
- 教程、明确操作路径、资源交付 →
- 读取目标 lane 的
.cheat-state.json，确认content_form一致。观点视频固定使用工作流项目的calibration/opinion-video/；无 lane 时先在该目录用上游cheat-init创建，不复用不匹配的旧目录。 - 先运行
node scripts/verify-blind-rubric.mjs calibration/<content_form>。对当前01-口播稿.md运行上游cheat-score：主评分器可用完整rubric_notes.md解析公式，但隔离子评分器只可读取已验证的blind-rubric.md与本期稿。任何 contamination warning 都必须丢弃并重启全新隔离评分。完整输出必须保留原始隔离 JSON（完整 script hash、rubric version、各维分数/理由、input_status、self_check、refusal）；登记脚本会复核它，不能只登记 Markdown 的 SHA。把无告警输出保存到 lane 内，再运行node scripts/register-cheat-score-evidence.mjs ...。- lane 尚无真实发布样本时（
calibration_samples=0），状态必须标为枚举值COLD_START_DIAGNOSTIC，不是白话“冷启动诊断”；register/prepare 脚本与 bianpai gate 按该枚举值机械判别，不得声称已验证质量或流量，也不得为了提高 composite 覆盖原版模板、口头禅、场景、节奏或金句。评分建议只有指出具体理解/交付问题时才进入改稿候选。
- lane 尚无真实发布样本时（
- 进入⑤：固定调用
dbs-script-flow、dbs-resonate、dbs-ai-check和laohan-shencha CONTENT_CLAIMS；只有首 5 秒被判弱时调用dbs-hook。dbs-ai-check 只诊断具体命中句，不自动改写；dbs-hook 一旦触发，完整使用上游三种方法，每种3—5条、总计10—15条差异候选。严禁补造作者经历、采访人数、数据、结果或案例，缺证据必须写[需真实证据]。报告完成后运行node scripts/stamp-episode-script-hash.mjs episodes/<slug> 04-深扫报告.md 04-事实核验.md，两份⑤报告才会写入同一个 script_hash。仅 hash 匹配不足：③必须没有明确未解决高风险；⑤深扫必须review_status: CLEAR与unresolved_issue_count: 0，事实核验必须fact_check_status: CLEAR、contradicted_count: 0、unverifiable_count: 0，否则不得 prepare 或登记最终盲预测。⑤还必须输出04-事实主张.json：任何 PROOF beat 都必须引用当前稿/事实报告绑定的SUPPORTED claim_id与同源 evidence，INFERRED（即使带inference_note）不得进入 PROOF beat，来源链接本身不构成事实证明。注意：当前 stamp 脚本只绑定两份⑤报告 hash，04-事实主张.json的 PROOF beat 边界属描述层约束，本次不改脚本。 - ⑤报告与当前稿 hash 一致后，先运行
node scripts/prepare-cheat-prediction-input.mjs episodes/<slug> calibration/<content_form>。把输出的 lane snapshot 作为上游cheat-predict的唯一输入，并在预测 metadata 写**Script**: <snapshot path>与**Script Hash**: <sha256>。然后运行node scripts/register-cheat-prediction.mjs episodes/<slug> calibration/<content_form> predictions/<file>.md v1。它验证预测晚于快照、metadata 与快照一致，才写03-预测证据.json并标RECORDED。任何改稿都会使 hash 失效；必须重跑⑤、prepare 新快照并登记新的 prediction revision，不能改写旧预测。 03-校准报告.md的开头必须是：
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.
- 16d ago First seen · 61 lines · 99 tokens per session scan A d849e1aa6cbd
laohan-jiaozhun is a skill published in the GitHub repository hanzhcn/laohan-skills (13 stars, last pushed yesterday), licensed MIT. It adds 99 tokens to every session and 2,038 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 2026-10-01.
Other skills, from other repositories
douyin-upload
A workflow for using the sau command-line tool to log in to Douyin, the Chinese short-video platform, check account cookies, and publish videos or image-and-text posts.
kuaishou-upload
A workflow for using the sau command-line tool to log in to Kuaishou, a Chinese social media platform, check account cookies, and publish videos or image-and-text posts.
xiaohongshu-upload
A workflow for using the sau command-line tool to log in to Xiaohongshu, a Chinese social media platform, check account cookies, and publish videos or image-and-text posts.
officecli-word-form
Use this skill to create fillable Word forms (.docx) with real Content Controls (SDT) + legacy FormField checkboxes + MERGEFIELD mail-merge placeholders + document protection. Trigger on: 'fillable form', 'form fields', 'content controls', 'SDT', 'word form', 'fill in', 'only editable fields', 'protect document'…
officecli-data-dashboard
Use this skill to build a multi-element Excel dashboard — Dashboard sheet on open, multiple formula-driven KPI cards, multiple charts, sparklines, and conditional formatting — from CSV or tabular input. Trigger on: 'dashboard', 'KPI dashboard', 'analytics dashboard', 'executive dashboard', 'metrics dashboard', 'CSV to…
invest
First-time fork users: run the invest-setup skill first to initialize. The main flow lives in this skill (the AI agent uses the CLI/MCP to view the portfolio / run the committee / replay decision history). The Web GUI has been retired (2026-07) — every capability is exposed via CLI subcommands / MCP tools. The backend…