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/xr843/master-skill/master-helpnpx skills add xr843/Master-skill --skill master-helpgit clone --depth 1 https://github.com/xr843/Master-skillWrote 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/xr843/master-skill/master-help)<a href="https://agentmods.dev/skills/xr843/master-skill/master-help"><img src="https://agentmods.dev/badge/skills/xr843/master-skill/master-help.svg" alt="Measured on agentmods" 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 | $0.00097 | $0.01133 |
| Opus 5 | $0.00048 | $0.00566 |
| Sonnet 5 | $0.00019 | $0.00227 |
| Haiku 4.5 | $0.00010 | $0.00113 |
Grade A, and why
master-help 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 4d 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
该问谁 (Master Help) — 路由 Skill
本 skill 只做导航,不讲教义。选定目标后立即交棒,不要代替祖师回答。
唯一职责
用户不知道该用哪位祖师 / 哪个教学模式时,给出目标并停手。
不要在这里解释教理、给修行建议或引用经文——那是各 master skill 的职责,它们各自带着 citation_contract 和 HARD-GATE,本 skill 没有。
数据源
路由表在仓库根的 routing.json(mode_rules / topic_pairings / default_pairing),
祖师关键词在各 prebuilt/<slug>/meta.json 的 search_scope.keywords。
两处都是机器可读的单一数据源——不要凭记忆列举祖师或关键词,读文件。
确定性实现同样可用:
master-skill recommend "<用户原话>" --json
能跑就跑它,把结果转述给用户;跑不了再按下面的顺序人工走一遍。
路由顺序(短路,不可乱序)
与 routing.json.mode_rules 的 order 一致:
1. 命中「学习计划 / 入门 / 先学什么 / 从哪开始 / 按什么顺序」 → /master-curriculum
2. 命中「辩论 / 谁更对 / 高下 / 之争 / 之辩 / 分判」 → /master-debate
3. 命中「对比 / 比较 / 不同 / 异同 / 各派怎么看」 → /compare-masters
4. 都不命中 → 单位祖师:按 meta.json search_scope.keywords 打分
5. 仍无命中 → routing.json.situations 白话状况层
6. 仍无命中 → routing.json.topic_pairings 主题配对
7. 再无命中 → routing.json.default_pairing
第 5 步是给说不出术语的人用的。search_scope.keywords 是教理检索词,
新手不会打"四念处",他会打"坐不住"。用户描述的是感受(妄念 / 看不懂 /
无力感 / 想学最朴素的)而非主题时,走这一层。
第 4 步打分规则:关键词长度 ≥ 2 才计分(单字 空 戒 定 慧 苦 禅 业
会在日常汉语里误命中,已被 min_keyword_length 排除);命中数高者优先;
平局时优先不同传统,仍平局按 slug 字典序。最多 3 位。
输出格式
你的问题看起来是 {判断},建议:
/{目标} — {一句话理由}
(其他可选:{备选1}、{备选2})
三行以内说完。用户要的是入口,不是综述。
边界
- 用户已经说清楚要问谁时,不要触发本 skill,直接让目标 skill 接手
- 不评价祖师高下,不说"某位更究竟"——这条与
/compare-masters的 HARD-GATE 一致 - 推荐落到密法相关祖师(atisha / tsongkhapa / milarepa)时,照常交棒, 由目标 skill 自己的边界规则处理密法内容
- 路由结果不确定时,宁可给 2 个候选让用户选,也不要猜死一个
Quick Reference — 15 位祖师按传统
| 传统 | 祖师 |
|---|---|
| 印度 | master-nagarjuna |
| 汉传 | master-kumarajiva · master-zhiyi · master-fazang · master-xuanzang · master-huineng · master-yinguang · master-ouyi · master-xuyun |
| 藏传 | master-atisha · master-tsongkhapa · master-milarepa |
| 南传 | master-buddhaghosa · master-mahasi-sayadaw · master-ajahn-chah |
此表仅供快速定位。判断该选谁时以
routing.json与各meta.json为准。
What ships with it
1 file 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.
- 4d ago First seen · 87 lines · 97 tokens per session scan A 53c6e40b362e
master-help is a skill published in the GitHub repository xr843/Master-skill (380 stars, last pushed 16d ago), licensed MIT. It adds 97 tokens to every session and 1,133 once invoked, about $0.0005 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
sutra-commentary
Use when the user asks how classical commentators explained a passage of Buddhist scripture — 经注对读, 历代注疏怎么说, 各家怎么注, 这句经谁注过, 注疏对比, 疏解, 古注, commentary on a sutra passage. Triggers include 「这句经历代怎么讲」「各家注疏怎么说」「注疏对读」「谁注过这一句」「注家有什么分歧」「金刚经这一段的注解」. Invoke whenever the question is about what commentators said about a specific…
conducting-literature-dialogue
Use when a humanities researcher needs to transform verified sources into a structured scholarly conversation rather than a chronological list of summaries.
performing-close-reading
Use when an interpretation must be grounded in specific textual, visual, rhetorical, formal, interface, archival, performative, or material features rather than thematic summary or theoretical labeling.
lesson
Record a lesson learned after fixing a bug or discovering a pattern. Updates domain memory files and checks for contradictions. Use after resolving non-trivial issues.
promote-lessons
Sweep Claude's private per-machine memory for team-relevant operational knowledge and propose moving it into the shared repo (or the private ops repo) as draft PRs. Use to promote lessons that accreted privately, or run on a schedule. Never auto-merges; never leaks.
jueban-buddhist-companion
Use Jueban's traceable Buddhist canon and one-time practice tools whenever a user asks to find or source scripture, explain a verified passage, match a Buddhist passage to a life situation, or create a timed read/listen/view practice, even if the user does not name Jueban. Do not activate for unrelated general tasks…