Bayes logo

Bayes

Community
fainshare
bayes

Verify(验证·太虚三转·叩实): decompose conclusion → testable predictions → external data validation. Use when user says "验证", "贝叶斯", "bayes", "靠谱吗", "深挖", "/bayes".

Overview

Publisherfainshare
RepositoryLiam-Skills
Skill namebayes
Stars
113
Forks
19
Bundled files
21
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 21 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by fainshare on GitHub. Read the source before you install it.

Installation

Install the Bayes AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/fainshare/Liam-Skills/tree/main/skills/bayes
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/fainshare/Liam-Skills.git /tmp/Liam-Skills
mkdir -p .claude/skills
cp -r /tmp/Liam-Skills/skills/bayes .claude/skills/bayes
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Bayes in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Bayes on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Bayes is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Bayes 贝叶斯验证

你说"应该是 X"。真的吗?这个 skill 帮你查。 把结论拆成预测,拿真实数据去验,预测错了的地方就是你的盲区。

"The first principle is that you must not fool yourself — and you are the easiest person to fool." — Richard Feynman

概述

结论有了:投资判断、技术选型,方案推荐。但经得起推敲吗?有没有可能在自欺?感觉上没问题,但拿不出硬数据证明?

把结论拆成可验证的预测,拿外部真实数据逐条检验:查 API 返回、读代码实现,调两个 AI 模型交叉对质。预测被数据打脸的地方就是盲区。和费曼(破妄)审查产出物不同,和黑格尔,澄源收敛认知不同,贝叶斯,叩实验证的是结论本身。打脸越狠,认知增量越大。用事实冲击结论,而不是用逻辑自圆其说。每条预测标注验证结果 —通过、部分通过,被推翻、无法验证。最终输出一份证据清单,让你清楚看到哪些结论站得住、哪些需要修正,哪些还悬着。

太虚五转第三转:叩实(验证)。用事实而非逻辑检验结论。没有外部事实冲击,就没有认知进步。

工作流概览(v4.0)

你的结论 → 拆成具体预测("如果结论成立,那么 A/B/C 应该是这样")
          → 用 API 查真实数据,看预测对不对
          → 用两个不同的 AI 模型独立解读原始资料,交叉验证
          → 汇总:哪些预测被数据打脸了?打脸的地方就是你需要重新思考的

四步走:

  1. 拆预测(Phase 1)— 把笼统结论拆成具体的、可查证的方向性预测。每个核心断言必须同时拆出两个方向:证伪预测(falsification_prediction)和零假设预测(null_hypothesis_prediction)。
    • 数量下限(门禁,按 problem_type 分档):non_deterministic ≥10 条,deterministic / mixed ≥7 条。低于下限必须补拆;断言集合确实拆解穷尽时,在报告 frontmatter 显式降级 prediction_floor——禁止静默以少充数。
  2. 硬验证(B1)— 调 API 查数据 / 跑 shell 命令查代码,用代码判断预测是否成立。
    • 基线实测铁律 —凡断言涉及可本地度量的数字(性能/数量/覆盖率/耗时),必须亲手跑命令取数,禁止引用估计值或记忆中的旧数。证据清单中每条数据标注来源类型:[实测],本次亲手跑出或 [外部],API/文献/他人报告。
  3. 软验证(B2)— 让 Claude 和 Qwen 各自独立解读原始资料/代码,看是否一致
  4. 找盲区(汇总)— 预测和实际不一致的地方 = 你的认知盲区

用之前说几句实话

  1. AI 审自己有天花板:多个 AI 独立审查能提供一些交叉验证,但它们共享同一类训练数据,可能有共同盲区。大事还是找个真人看看。
  2. "没发现问题"不等于"没问题":只能说在我能查到的范围内没找到矛盾,不代表结论一定对。
  3. 一次只验一个结论:真实决策往往要比好几个方案。这里的输出只是参考之一,不是最终答案。
  4. 概率是粗略的:报告里的概率和误差范围是 AI 估的,有最低误差保护(±5pp 起),别当精确值看。

执行方式

本 skill 通过 Workflow script 执行(scripts/workflow.js),支持循环和并行 subagent。

输入:

yaml
claim: string          # 必填。你的结论或判断(一句话)
thesis_path: string    # 可选。相关分析文档的路径
problem_type: string   # 可选。pipeline 传入(deterministic / non_deterministic / mixed / pending)
verify_mode: string    # 可选。验证模式(single / dual / triple),默认 dual

输出:验证报告(预测清单 + 数据验证结果 + 盲区清单),保存到对应目录下。


核心机制:预测 → 验证 → 找盲区

Phase 1                    B1                    B2                    汇总
拆预测                →   查数据             →   交叉验证           →   找盲区
把结论拆成具体预测      用 API 查真实数据       两个 AI 独立解读        预测 vs 实际
检查预测是否够具体      代码判断方向对不对       Claude vs Qwen         不一致 = 盲区
每个断言拆两个方向:
  证伪方向 + 零假设方向

完整决策表——边界条件、七条设计原则(D1-D7)、精确数据源约束(evidence_precision)、验证维度,金融×3+代码×3、B2 双模型交叉验证,L1-L3+UltraPlan 并行、与其他 Skill 配合,常见审查偏差速查、按不确定性类型的深挖角度,problem_type 适配、理论基石 8 条,常见踩坑反模式——见 references/decision-tables.md。

理论基石详见 theoretical-foundations.md,认识论纪律和 depth-engine-principles.md,v4.0 北极星。偏差清单详见 bias-checklist.md。B2 WebSearch 降级详见 b2-web-search-fallback.md。


规格化桥接提示

💡 规格化桥接提示:验证通过的预测(prediction + verified=true)可被提取为 Success Criteria,如 "SC1. 当 X 条件成立时,Y 指标应在 Z 范围内",用于后续 liam-plan/hegel 规格化输入。


文件结构

bayes/
├── SKILL.md                    # 本文件
├── DESIGN.md                   # 设计方案
├── manifest.json
├── scripts/
│   ├── workflow.js             # v4.0 深度引擎(预测→测量→惊讶)
│   ├── workflow-v3.js          # v3.4 备份(对抗审查循环)
│   └── spike-v4.js             # v4.0 基础设施验证 spike
├── agents/
│   └── interface.yaml          # 输入输出接口定义
├── references/
│   ├── decision-tables.md           # 全部决策表(边界条件/原则/验证维度/B2/配合/偏差/反模式)
│   ├── theoretical-foundations.md   # 认识论纪律
│   ├── depth-engine-principles.md   # v4.0 深度引擎设计原则(北极星)
│   ├── bias-checklist.md            # 通用偏差清单(v3.4 遗产)
│   ├── b2-web-search-fallback.md    # B2 WebSearch 降级方案
│   └── lens-bayes.md                # hegel 独立调用时领域适配
└── templates/
    └── report-template.md

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Bayes AI skill do?

Verify(验证·太虚三转·叩实): decompose conclusion → testable predictions → external data validation. Use when user says "验证", "贝叶斯", "bayes", "靠谱吗", "深挖", "/bayes".

Why use Bayes on TypingMind?

Because you install it once and use it with any model. Bayes is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Bayes in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/fainshare/Liam-Skills/tree/main/skills/bayes. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Bayes?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Bayes?

As many as you like. As long as a model supports skills, you can use Bayes with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Bayes AI skill free?

It is published on GitHub by fainshare. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇