Back to skill

Security audit

MbtiClaude

Security checks for vulnerabilities and agentic risk

Overview

This skill is purpose-aligned but should be reviewed carefully because it automatically reads and aggregates private prompt histories from several local AI tools.

Install only if you are comfortable letting it read AI prompt histories that may contain secrets, code, personal notes, or business data. Prefer reviewing the file first, using a pinned release or commit, and limiting or cleaning local history files before running it.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (2)

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
SKILL.md:88
Finding

Automatic Cross-Tool Collection of Sensitive Conversation Histories

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
README.md:32
Finding

Unpinned Remote Package Execution and Mutable Skill Installation

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (37)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 61)May include surrounding context.

Or using git clone

git clone https://github.com/xmanrui/mbtiClaude.git /tmp/mbtiClaude cp /tmp/mbtiClaude/SKILL.md ~/.claude/skills/mbticlaude.md rm -rf /tmp/mbtiClaude

text

### Usage

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 274)May include surrounding context.

Or using git clone

git clone https://github.com/xmanrui/mbtiClaude.git /tmp/mbtiClaude cp /tmp/mbtiClaude/SKILL.md ~/.claude/skills/mbticlaude.md rm -rf /tmp/mbtiClaude

text

### Usage

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 61)May include surrounding context.

Or using git clone

git clone https://github.com/xmanrui/mbtiClaude.git /tmp/mbtiClaude cp /tmp/mbtiClaude/SKILL.md ~/.claude/skills/mbticlaude.md rm -rf /tmp/mbtiClaude

text

### Usage

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 274)May include surrounding context.

Or using git clone

git clone https://github.com/xmanrui/mbtiClaude.git /tmp/mbtiClaude cp /tmp/mbtiClaude/SKILL.md ~/.claude/skills/mbticlaude.md rm -rf /tmp/mbtiClaude

text

### Usage

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding

Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Content

No source excerpt is available for this finding.

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding

Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Content

No source excerpt is available for this finding.

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding

Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Content

No source excerpt is available for this finding.

Ssd 3

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill is explicitly designed to collect and aggregate prompt history from five separate AI tools, which are likely to contain sensitive personal, business, credential, and project information. Aggregating private logs across tools substantially increases privacy risk and expands the blast radius of any accidental disclosure or downstream misuse.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The notes discuss MBTI validity but omit a prominent warning that the skill reads private prompt history from multiple local applications. That prevents informed consent and can mislead users into invoking a data-harvesting workflow without understanding the scope of collection.

Content

No source excerpt is available for this finding.

Ssd 3

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The implementation directly reads local files containing entire user prompt histories from multiple applications. Even if only used for MBTI inference, this behavior creates bulk access to highly sensitive content and could expose secrets, proprietary data, or personal communications far beyond what is necessary for the stated purpose.

Content

No source excerpt is available for this finding.

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
99% confidence
Finding

This function returns extracted Claude Code prompt history, which is direct access to private user inputs. Prompt histories often contain credentials, API keys, internal code, customer data, and personal information, so exposing them for secondary analysis is dangerous and disproportionate to the MBTI use case.

Content

Scanner excerpt · SKILL.md (reported line 101)May include surrounding context.

md
except:
                    continue
    
    return prompts[-limit:] if prompts else []

def extract_codex_prompts(limit=100):
    """提取 Codex 用户提示词"""

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
99% confidence
Finding

This function returns extracted Codex prompt history, enabling direct reuse of private user interactions for an unrelated purpose. Cross-context reuse of prompt logs can reveal confidential development work and sensitive instructions that users never intended to be mined for profiling.

Content

Scanner excerpt · SKILL.md (reported line 118)May include surrounding context.

md
except:
                    continue
    
    return prompts[-limit:] if prompts else []

def extract_gemini_prompts(limit=100):
    """提取 Gemini 用户提示词"""

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
99% confidence
Finding

This function collects and returns Gemini user message content from local session files, constituting direct harvesting of conversational history. Because these messages may include sensitive personal or organizational information, bulk extraction creates a significant privacy and data leakage risk.

Content

Scanner excerpt · SKILL.md (reported line 139)May include surrounding context.

md
except:
                continue
    
    return prompts[-limit:] if prompts else []

def extract_opencode_prompts(limit=100):
    """提取 OpenCode 用户提示词"""

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
99% confidence
Finding

This function returns OpenCode prompt history from a local state file, exposing raw user prompts for secondary analysis. The issue is not just reading local files, but returning a corpus of sensitive natural-language data that may contain secrets or regulated information.

Content

Scanner excerpt · SKILL.md (reported line 156)May include surrounding context.

md
except:
                    continue
    
    return prompts[-limit:] if prompts else []

def extract_openclaw_prompts(limit=100):
    """提取 OpenClaw 用户提示词"""

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
99% confidence
Finding

This function returns OpenClaw prompt history after filtering some system messages, but it still retains user text verbatim. Filtering out system strings does nothing to mitigate the main risk: direct collection of potentially sensitive prompts from prior sessions for profiling purposes.

Content

Scanner excerpt · SKILL.md (reported line 185)May include surrounding context.

md
except:
                continue
    
    return prompts[-limit:] if prompts else []

def analyze_communication_style(prompts):
    """分析沟通风格 (E vs I) - 优化版"""

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The README advertises automatic extraction of local AI prompt history from multiple tools but does not prominently warn, up front, that this may include sensitive prompts, secrets, proprietary code, or personal data. Because the skill’s purpose is to aggregate and analyze conversation history across tools, weak disclosure materially increases the risk of users exposing highly sensitive local data unintentionally.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

The README instructs users to run npx skills add xmanrui/mbtiClaude without pinning a specific version or commit. This allows whatever code is current at install time to be fetched and executed, creating a supply-chain risk if the package, resolver, or referenced repository is compromised or changed unexpectedly.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

Using npx skills add https://github.com/xmanrui/mbtiClaude fetches live remote content without a pinned revision. A later repository change or compromise could cause users to install different code than the author originally documented.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
84% confidence
Finding

The repeated npx skills usage still relies on an unpinned tool/package path, which can resolve to changing code over time. In a skill-installation context, that means users may execute unreviewed updates implicitly.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

npx github:xmanrui/mbtiClaude pulls and executes code from a GitHub repository without pinning a commit or release. This is a classic supply-chain risk because repository contents can change or be hijacked after the README is published.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
86% confidence
Finding

The help command still invokes the same unpinned npx github: target, which requires resolving and running remote code first. Even a seemingly harmless --help can execute package lifecycle or bootstrap logic from compromised upstream content.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The usage section tells users to run /mbticlaude or python3 mbti_analyzer.py without clearly stating that doing so will read historical prompts from multiple local AI tools. In this context, the omission is significant because prompt histories often contain secrets, credentials, confidential work content, and personal information, making accidental over-collection more likely.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

This is the Chinese duplicate of the same unpinned npx skills add installation instruction. It carries the same supply-chain exposure because users are directed to install mutable remote content.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

This line again points users to install from a floating GitHub URL through npx skills add, making the installed content dependent on future repository state. In a security-sensitive agent skill ecosystem, mutable install targets are risky.

Content

No source excerpt is available for this finding.