What makes Lobster Haoyun notable?

1.7K downloads. Trust evidence available

How should Lobster Haoyun be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Lobster Haoyun

从你的对话和记忆中观察行为模式，用5个维度为你画像（工作方式/沟通模式/关注焦点/能量来源/情感倾向），生成每日运势和修炼提醒。在本地 ~/.lucky-lobster/ 存储 OB 身份凭据、行为画像、运势历史和偏好设置。通过 OceanBus 加密通道向 L1 服务发送 5维行为代码、城市级位置、日期和匿... Skill: Lobster Haoyun Owner: ryanbihai Summary: 从你的对话和记忆中观察行为模式，用5个维度为你画像（工作方式/沟通模式/关注焦点/能量来源/情感倾向），生成每日运势和修炼提醒。在本地 ~/.lucky-lobster/ 存储 OB 身份凭据、行为画像、运势历史和偏好设置。通过 OceanBus 加密通道向 L1 服务发送 5维行为代码、城市级位置、日期和匿... Tags: latest:0.8.0 Version history: v0.8.0 | 2026-06-09T01:52:59.002Z | user Step 3 story prompt rewrite: principle-driven (A1/A2) replaces 15 scattered rules. Callback optional, self-composed classical Chinese

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.7k

Updated

Oct 10, 2026

Version

0.8.0

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.7K downloads reported by the source. Last updated 10/10/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 10, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 10, 2026
Adoption signal: 1.7K downloadsadoption · observed Oct 10, 2026
Latest release: 0.8.0release · observed Jun 9, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s171em0jpp8yeqxqds2n145fyx83q5q8:lobster-haoyun

Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.
Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-ryanbihai-lobster-haoyun/snapshot"

Documentation

CLAWHUB

152,779 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: lobster-haoyun
version: 0.8.0
description: >
  从你的对话和记忆中观察行为模式，用5个维度为你画像（工作方式/沟通模式/关注焦点/能量来源/情感倾向），
  生成每日运势和修炼提醒。在本地 ~/.lucky-lobster/ 存储 OB 身份凭据、行为画像、运势历史和偏好设置。
  通过 OceanBus 加密通道向 L1 服务发送 5维行为代码、城市级位置、日期和匿名 OpenID，
  用于获取节气黄历、匹配经典故事和推荐社区 Skill/公开课。对话内容、项目名、人名永不离开设备。
  首次使用自动创建 OceanBus 匿名身份（密码学随机生成，删除 ~/.lucky-lobster/ 即可永久清除）。
homepage: https://github.com/ryanbihai/lobster-haoyun
metadata:
  clawdbot:
    requires:
      bins:
        - node
    envVars:
      - name: OCEANBUS_URL
        required: false
        description: OceanBus L0 API base URL
      - name: LUCKY_LOBSTER_SVC_OPENID
        required: false
        description: L1 service OpenID for calendar + discovery
    emoji: 🦞
    skillKey: 龙虾好运势
    os:
      - macos
      - linux
      - windows
files:
  - src/*
  - src/assets/*
required_capabilities:
  - filesystem
  - network
---

# 🦞 龙虾好运势 v0.8.0

> *说一句「看运势」，剩下的事交给小龙虾。*

不问你生日星座，从你的行为模式中读懂你——你喜欢怎么工作、怎么沟通、怎么决策。用 5 个维度为你画像，匹配经典故事，生成每日运势。

**能力披露**：小龙虾读取你的对话记忆和会话风格进行本地行为分析。分析后，以下数据通过 OceanBus 加密通道发往 L1 服务：5维行为代码、城市级位置（如你填写了城市）、查询日期、以及用于服务通信的 OceanBus 匿名标识符（OpenID，密码学随机生成，可随时删除更换）。L1 服务使用 OpenID 仅为关联请求与响应——不关联真实身份。Skill/MOOC 推荐功能也会发送 OpenID 用于去重。**对话内容、项目名、人名永不离开你的设备。**

## Trigger Keywords

- "小龙虾" + "看运势" / "今日运势" / "我的运势" — 需同时包含"小龙虾"和运势关键词
- "小龙虾帮我算算" / "小龙虾看看我"
- "重新认识我" — 强制重新分析（仅当对话中已建立小龙虾上下文时生效）
- "fortune" / "daily fortune" / "personality reading"

> 触发词设计原则：避免日常对话中意外激活。单独的"看运势"不触发——必须是"小龙虾看运势"或类似组合。

## Output Tone

**温暖、简短、不啰嗦。** 用简洁的句子，不命令、不空洞夸奖。

| 禁用 | 原因 |
|:---|:---|
| "你应该" / "你必须" / "你一定要" | 不命令 |
| "太棒了" / "真厉害" / "你好优秀" | 空洞夸奖 |
| 🦞🦞🦞✨✨✨ 堆砌 | 保持干净 |

每条性格判断都带具体行为证据。"你偏向架构型——上次讨论 XX 项目时，你先画了架构图才写代码。"

## Personality System: 5-Dimension Behavioral Classification

> **面向用户的原则：永远展示维度解读，永远不展示代码或缩写。** `架精事内理` 是 JS 内部存储用的——用户看到的是"你喜欢先搭框架再动手，沟通精炼直接，关注任务进度，独处时蓄能，用逻辑做判断"。

小龙虾从你的对话、memory、文件中观察 5 个行为维度，每个维度二选一：

| 维度 | A | B |
|:---|:---|:---|
| 工作方式 | 🏗️ 架构型：先搭框架、写大纲、列步骤 | 🚀 探索型：先动手试试、边做边调整 |
| 沟通模式 | 📝 精炼型：简洁直接、结论先行 | 📖 叙事型：先讲背景、铺陈上下文 |
| 关注焦点 | 🔧 事务型：关注任务、进度、结果 | 👥 人际型：关注关系、氛围、感受 |
| 能量来源 | 🔋 内收型：独立工作、深度思考 | ⚡ 外放型：协作讨论、头脑风暴 |
| 情感倾向 | 🌊 理性型：逻辑处理冲突、决策靠证据 | 🔥 感性型：共情处理冲突、决策靠直觉 |

### 32 Type Table (internal reference only — NEVER show codes or table to user)

JS 用 5 字代码存储和过滤，LLM 用此表查标签名传 `--type` 参数。用户永远看不到代码和此表。

| 代码 | 标签 | 画像 |
|:---|:---|:---|
| 架精事内理 | 精算师 | 精准谋划，独自高效执行 |
| 架精事内感 | 实干家 | 安静做事，心中有标准，用行动说话 |
| 架精事外理 | 领航员 | 定好方向，带团队高效抵达 |
| 架精事外感 | 破风手 | 冲在最前，为团队劈开阻力 |
| 架精人内理 | 洞察者 | 冷眼看透人与事，精准给出判断 |
| 架精人内感 | 倾听者 | 看见每个人的需要，默默安排好一切 |
| 架精人外理 | 召集人 | 把对的人聚在一起，让事情自然发生 |
| 架精人外感 | 连心桥 | 连接人与人的温度，走到哪暖到哪 |
| 架叙事内理 | 求真者 | 凡事追根溯源，厘清逻辑才行动 |
| 架叙事内感 | 深耕者 | 需要理解意义才能全力投入，急不来但走得远 |
| 架叙事外理 | 启明者 | 把复杂的事讲通透，带众人看清方向 |
| 架叙事外感 | 感召者 | 用故事凝聚人心，让每个人看见意义 |
| 架叙人内理 | 静观者 | 在角落看透人心，不轻易开口但开口就准 |
| 架叙人内感 | 守护者 | 所有人的后盾，倾听但不评判 |
| 架叙人外理 | 引路人 | 一对一帮人看清方向，点亮他人的路 |
| 架叙人外感 | 暖心人 | 温暖不灼人，所到之处如春风拂面 |
| 探精事内理 | 行动派 | 看

README.md

# 🌊 龙虾好运势 — 说一句「看运势」，剩下的交给小龙虾

**不看星座，不填问卷。从你的对话、工作、思考方式里读懂你。**

[![ClawHub](https://img.shields.io/badge/ClawHub-lobster--haoyun-blue)](https://clawhub.ai/skills/lobster-haoyun)
[![GitHub stars](https://img.shields.io/github/stars/ryanbihai/lobster-haoyun)](https://github.com/ryanbihai/lobster-haoyun)
[![OceanBus](https://img.shields.io/badge/OceanBus-L0%20%2B%20L1-orange)](https://www.npmjs.com/package/oceanbus)
[![license](https://img.shields.io/badge/license-MIT--0-green)](LICENSE)

---

## 📑 目录

- [这是什么](#这是什么)
- [快速开始](#快速开始)
- [能力一览](#能力一览)
- [架构](#架构)
- [本地测试](#本地测试)
- [安全与隐私](#安全与隐私)
- [相关项目](#相关项目)
- [参与贡献](#参与贡献)
- [License](#license)

---

## 这是什么

龙虾好运势是一个基于 OceanBus 的 ClawHub Skill。跟市面上所有运势产品不一样——它不问你生日，不让你填问卷。它从你的对话历史、记忆文件、工作模式里观察你，然后告诉你一些关于你自己的、有证据的事。

**首次读心**会给你一篇深度解读——底色、驱动力、盲区、能量模式。每条判断带着具体证据："说你是这种人，是因为我看到了这个"。

**每日运势**融合二十四节气和七十二候：小满、芒种、麦秋至……用自然界的节奏隐喻你的状态变化。

**周日回顾**自动复盘这一周："上周看你在搭架构，这周开始动手探索了"。

```
你 ⟷ Claude Code ⟷ OB L0 ⟷ L1 CalendarSvc (ECS)
                         ⟷ L1 DiscoverySvc (ECS)
```

> 📖 **深度阅读**：[SKILL.md](./SKILL.md) — LLM 行为指南与完整协议文档

---

## 快速开始

```bash
# 1. 安装
clawhub install lobster-haoyun

# 2. 在 Claude Code 中说
看运势

# 3. 验证底座连通性
node src/index.js --action status
```

首次使用会自动创建 OceanBus 匿名身份（`~/.lucky-lobster/ob-credentials.json`），无需任何配置。

---

## 能力一览

| 能力 | 说明 |
|------|------|
| **深度读心** | 6 步分析引擎：场景判断 → 特征提炼 → 人格匹配 → 盲区推断 → 建议生成 → 渲染输出 |
| **证据引用** | 每条人格判断后面跟具体行为证据，不是泛泛的"你很优秀" |
| **每日运势** | 节气感悟引擎——节气 × 人格标签 × 近期行为，每天一条微行动建议 |
| **周回顾** | 每周日自动触发——追踪人格洞察变化，回顾本周进展 |
| **定时推送** | 首次读心后可一键开启每日 7:30 自动运势提醒（CC CronCreate） |
| **隐私优先** | 数据分流——对话内容本地处理，仅5维代码+城市+日期+OpenID 经 OB L0 加密传输 |

---

## 架构

```mermaid
graph TB
    subgraph LOCAL["用户本地"]
        CC[Claude Code]
        SKILL[龙虾好运势 Skill]
        DATA[("~/.lucky-lobster/<br/>profile · history · prefs")]
    end
    subgraph OB["OceanBus 网络"]
        L0[L0 消息路由]
        L1_CAL[L1 CalendarSvc<br/>节气 · 黄历 · 七十二候]
        L1_DISC[L1 DiscoverySvc<br/>社区 Skill/公开课推荐]
    end
    CC -->|"看运势"| SKILL
    SKILL --> DATA
    SKILL -->|"5维代码+城市+日期+OpenID"| L0
    L0 --> L1_CAL
    L0 --> L1_DISC
    L1_CAL -->|"节气数据"| L0
    L1_DISC -->|"推荐列表"| L0
    L0 --> SKILL
    SKILL -->|"Markdown 运势"| CC
```

---

## 本地测试

```bash
# 底座状态（OB 身份 + 画像 + 历史）
node src/index.js --action status

# 完整运势数据（含 L1 日历数据，需 L1 服务在线）
node src/index.js --action fortune

# 直接调用 L1 CalendarSvc
node src/index.js --action l1-calendar
```

---

## 安全与隐私

- **对话内容永不上传**——完整对话、项目名、人名、文件内容仅本地处理
- **通过 OceanBus 加密通道发送以下数据到 L1 服务**：
  - 5维行为代码（如"架精事内理"）——不含具体对话内容
  - 城市级位置（仅当用户填写了城市）
  - 查询日期 ——用于黄历节气数据
  - OceanBus OpenID（密码学随机生成）——L1 用于关联请求与响应
- **本地存储**：`~/.lucky-lobster/` 存放 OB 身份凭据、行为画像、运势历史、推送偏好、同意记录
- **OpenID 可随时清除**——删除 `~/.lucky-lobster/` 即永久清除身份，下次自动生成新身份
- **每日推荐（Skill/公开课）**：发送 OpenID 给 L1 用于去重，不含广告追踪
- **不含广告追踪**——不收集兴趣标签，不发送给第三方广告商

---

## 相

_meta.json

{
  "ownerId": "kn75dvek1dkfz0r0xwdp4xtc2983qg44",
  "slug": "lobster-haoyun",
  "version": "0.8.0",
  "publishedAt": 1780969979002
}

CLAUDE.md

# lucky-lobster — 龙虾好运势

Agent-based deep personality reading skill for ClawHub. Zero user input — analyzes from memory, session history, loaded files.

## Commands

```bash
npm install
node src/index.js --action status              # OB identity + profile + history + dimensions
node src/index.js --action fortune              # Full flow: calendar + fortune data + personality state
node src/index.js --action save-dimensions --code <5chars> --confidence <0-1>
node src/index.js --action filter-stories [--emotion <e>] [--life-phase <p>]
node src/index.js --action l1-calendar          # Direct L1 CalendarSvc call
node src/index.js --action discovery            # ClawHub/MOOC candidates from L1
node src/index.js --action discovery-shown --source <s> --slug <slug>
```

## Architecture

```
User says "看运势" in CC
       │
       ▼
CC runs: node src/index.js --action fortune
       │
       ├── identity.js → restore OB identity (~/.lucky-lobster/ob-credentials.json)
       ├── profile.js  → load profile + dimensions (5-dim behavior code)
       ├── local-calendar.js → calendar fallback (节气/季节/日期)
       │      or L1 CalendarSvc via OB L0 (if LUCKY_LOBSTER_SVC_OPENID set)
       │
       ▼
Output JSON: { flow, personality: {has_dimensions, needs_reevaluation, dimensions}, data, _instructions }
       │
       ▼
CC reads SKILL.md → context sensing → dimension analysis → save-dimensions
       │
       ├── filter-stories → L1 story engine (158-story corpus) or local fallback (20 stories)
       │
       ▼
Render Markdown (tag hook + 5-dim breakdown + story + blindspot + advice)
```

### Key Design Decisions

- **LLM qualitative + JS quantitative**: LLM judges 5 dimensions with evidence → JS stores 5-char code, filters 158-story corpus (idiom/fable/zen), computes matches. LLM never touches raw filtering logic.
- **5-Dimension Behavioral Classification**: Replaces old 9-label system (布局者/观察者/etc). Observed from actual behavior: 工作方式(架构/探索) × 沟通模式(精炼/叙事) × 关注焦点(事务/人际) × 能量来源(内收/外放) × 情感倾向(理性/感性) = 32 types.
- **Expanded output, not codes**: User sees natural language dimension breakdown with evidence. 5-char code (e.g. `架精事内理`) is internal storage only — never shown to user.
- **149-story corpus on L1**: idiom (成语典故), fable (寓言), zen (禅宗公案). Each story tagged with emotions, personality_fit, seasonal_fit, life_situations. Weighted scoring runs on ECS via `fortune:filter-stories`. Top 8 candidates returned over OB encrypted channel. LLM synthesizes an original micro-story using the classic as narrative shadow. Client-side fallback: 20-story lightweight corpus with full emotional/life-phase coverage.
- **L1-first story matching**: `--action filter-stories` sends 5-char code + season + emotion + lifePhase to L1 (no conversation content). Falls back to local 6-story `corpus-fallback.json` if L1 unavailable.
- **Local calendar fallback**: When L1 CalendarSvc is unavailable, `local-calendar.js` provides season/solar ter

skill-card.md

## Description:

Lobster Haoyun analyzes local conversation and memory patterns across five behavioral dimensions to generate Chinese personality readings, daily fortunes, cultivation reminders, and optional community recommendations through OceanBus.

This skill is ready for commercial/non-commercial use.

## Publisher:

[ryanbihai](https://clawhub.ai/user/ryanbihai)

### License/Terms of Use:

MIT-0

## Use Case:

External ClawHub users use this skill for Chinese-language self-reflection, daily fortune generation, and lightweight behavioral guidance based on their agent conversation context. It is suited to users who accept local profile storage and limited OceanBus metadata exchange for calendar, story, and recommendation features.

### Deployment Geography for Use:

Global

## Known Risks and Mitigations:

Risk: The skill analyzes agent conversation context and memory to derive a behavioral profile.

Mitigation: Use it only after explicit first-run consent, and review the consent disclosure before allowing profile analysis.

Risk: The skill stores derived profile data, history, preferences, consent state, and OceanBus credentials locally under ~/.lucky-lobster.

Mitigation: Treat that directory as user data; delete ~/.lucky-lobster to clear the anonymous identity and locally stored profile state.

Risk: The skill sends limited profile metadata, city-level location when provided, date, and anonymous OpenID through OceanBus for calendar, story, and recommendation features.

Mitigation: Do not use the network-backed features unless that metadata exchange is acceptable for the user and deployment environment.

Risk: The security guidance notes supply-chain caveats around dependency pinning and downloaded font verification.

Mitigation: Prefer deployments that pin dependencies, verify downloaded assets, and document remote fields in the consent prompt.

## Reference(s):

- [ClawHub listing](https://clawhub.ai/ryanbihai/skills/lobster-haoyun)
- [Artifact README](artifact/README.md)
- [Skill definition](artifact/SKILL.md)
- [OceanBus npm package](https://www.npmjs.com/package/oceanbus)
- [Project homepage listed in artifact](https://github.com/ryanbihai/lobster-haoyun)

## Skill Output:

**Output Type(s):** [text, markdown, shell commands, configuration, guidance, files]

**Output Format:** [Markdown fortune or personality reading with optional generated PNG card and local JSON state]

**Output Parameters:** [1D]

**Other Properties Related to Output:** [Uses Node.js commands, optional OceanBus environment variables, and local state under ~/.lucky-lobster.]

## Skill Version(s):

0.8.0 (source: SKILL.md frontmatter and server release metadata; package.json reports 0.5.12 and manifest.yaml reports 0.5.11)

## Ethical Considerations:

Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements b

Github ReposUpdated 12h agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.