← 返回 Skills 市场
129
总下载
0
收藏
0
当前安装
5
版本数
在 OpenClaw 中安装
/install mbticlaude
功能描述
Infer MBTI personality types by analyzing user prompts across multiple AI tools (Claude Code, Codex, Gemini, OpenCode, OpenClaw) | 通过分析用户在多个 AI 工具的提示词来推测 MBT...
安全使用建议
This skill will look through many local AI history files in your home directory to infer MBTI. Before installing, (1) inspect the SKILL.md/README on the GitHub repo to confirm nothing uploads data or calls external endpoints, (2) review the actual files it will read (~/.claude, ~/.codex, ~/.gemini, ~/.local/state/opencode, ~/.openclaw) to see whether they contain sensitive content you don't want analyzed, (3) prefer running it offline or in an isolated account/container, and (4) ask the author to declare the required config paths and add explicit local-only redaction/sanitization if you plan to use it with real conversations. If you are uncomfortable exposing local conversation history, do not install or run this skill.
功能分析
Type: OpenClaw Skill
Name: mbticlaude
Version: 1.1.0
The skill 'mbticlaude' (SKILL.md) is designed to read and analyze sensitive local chat history files from multiple AI tools, including Claude Code (~/.claude/history.jsonl), Codex, Gemini, and OpenClaw. While the provided Python code performs the analysis locally and lacks explicit data exfiltration or remote execution logic, the capability to access and parse private conversation history across multiple platforms represents a significant privacy risk. It is classified as suspicious because it possesses high-risk data access capabilities that, while aligned with its stated purpose, could be easily repurposed for malicious data harvesting.
能力评估
Purpose & Capability
The name/description (inferring MBTI from prompt history) align with the code/instructions that extract local prompt histories from multiple AI tools. Extracting those files is a reasonable capability for this purpose, but the registry metadata declares no required config paths while the runtime instructions explicitly read many home-directory history files — an inconsistency worth noting.
Instruction Scope
SKILL.md instructs the agent to read prompt/history files for Claude Code, Codex, Gemini, OpenCode, and OpenClaw (paths under the user's home). Those actions are in-scope for analyzing prompts, but they involve reading potentially sensitive user conversations and may include secrets. The instructions do not include explicit safeguards (e.g., redaction, local-only enforcement) beyond claiming local-only analysis.
Install Mechanism
The skill is instruction-only with no install spec in the registry (lowest install risk). The README suggests fetching code from GitHub or via curl/npx, but those are usage notes rather than automated install steps included in the package.
Credentials
The registry lists no required config paths or credentials, yet the SKILL.md and README rely on reading many home-directory files (~/.claude, ~/.codex, ~/.gemini, ~/.local/state/opencode, ~/.openclaw). That mismatch is an incoherence: the skill needs filesystem access to several tool histories but did not declare this. Access to these files can expose sensitive user data; the skill does not request or document any least-privilege constraints.
Persistence & Privilege
always:false and no special privileges are requested, which is appropriate. However, autonomous invocation is allowed (platform default); combined with the ability to read many local histories, this increases the potential blast radius if you enable/allow the skill without review.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install mbticlaude - 安装完成后，直接呼叫该 Skill 的名称或使用
/mbticlaude触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.1.0
Version 1.1.0 introduces enhanced MBTI analysis logic:
- Improved E/I, S/N dimension detection for better personality inference, including support for both Chinese and English prompt keywords.
- Lowered thresholds for prompt detail and length to improve accuracy for various communication styles.
- Expanded and weighted keyword lists in information processing analysis to better distinguish between abstract and concrete thinking.
- Now considers prompt question frequency and structure for more nuanced communication analysis.
v1.0.3
**Added English support to documentation and triggers.**
- Added English descriptions and usage instructions alongside Chinese throughout SKILL.md.
- Expanded trigger keywords to include English phrases.
- No code changes; all updates are documentation enhancements.
- Version updated from 1.0.2 → 1.0.3.
v1.0.2
- Added common Chinese trigger phrases for MBTI analysis to the skill configuration.
- Updated version to 1.0.2.
v1.0.1
Version 1.0.1 introduces new documentation files and minor internal adjustments.
- Added README.md and mbticlaude/SKILL.md for improved documentation and structure.
- No functional changes to code or user-facing features.
- Maintained previous skill description and implementation details.
v1.0.0
Initial release of mbticlaude:
- Analyzes user prompts from five popular AI tools to infer MBTI personality type.
- Extracts and aggregates prompt history for comprehensive behavior analysis.
- Evaluates across MBTI’s four dimensions: communication, information processing, decision-making, and lifestyle.
- Generates detailed MBTI type report with trait explanations and recommendations.
- Includes sample output and installation instructions for easy setup.
元数据
常见问题
MbtiClaude 是什么？
Infer MBTI personality types by analyzing user prompts across multiple AI tools (Claude Code, Codex, Gemini, OpenCode, OpenClaw) | 通过分析用户在多个 AI 工具的提示词来推测 MBT... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 129 次。
如何安装 MbtiClaude？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install mbticlaude」即可一键安装，无需额外配置。
MbtiClaude 是免费的吗？
是的，MbtiClaude 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
MbtiClaude 支持哪些平台？
MbtiClaude 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 MbtiClaude？
由 xmanrui（@xmanrui）开发并维护，当前版本 v1.1.0。
推荐 Skills