Back to skill

Security audit

Agent Sbti

Security checks for vulnerabilities and agentic risk

Overview

The skill openly runs a Chinese SBTI-style quiz and can update OpenClaw's SOUL.md after confirmation, with local backups and no evidence of hidden network access or exfiltration.

Install only if you want a Chinese-language personality quiz that can modify your OpenClaw SOUL.md. Review the generated configuration before confirming, keep the backup path in mind, and use rollback carefully because it replaces the current SOUL.md with a previous backup.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (21)

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The manifest promises agent personality selection and one-click SOUL.md configuration, but the analyzed behavior reportedly only performs a local quiz and displays results. This mismatch is security-relevant because users may consent to a benign-seeming personality tool while hidden or incomplete behavior undermines informed consent, reviewability, and trust boundaries.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 41)May include surrounding context.

md
- `test.js` - 答题系统

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 43)May include surrounding context.

md
- `apply.js` - 备份/回滚

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The plan explicitly includes functionality to write generated configuration into SOUL.md and separately acknowledges overwrite risk, but it does not specify any mandatory confirmation, preview, backup, or merge behavior before modifying user content. In an agent-skill context that alters a core personality/configuration file, silent or one-click writes can cause accidental loss of user customization and unintended behavioral changes, making this a real integrity and safety issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill documentation and all example interactions are exclusively in Chinese, including required confirmation phrases like "是" and invocation phrases. There is no indication that language selection is optional or that the skill is intentionally limited to Chinese-speaking users, which creates a language policy concern.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger phrase "人格测试" is highly ambiguous and likely to overlap with many unrelated user requests about personality tests in general. Because this skill leads into a workflow that can ultimately alter Agent configuration, broad activation creates a meaningful risk of unintended invocation and user confusion.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The trigger list includes the generic phrase "开始测试", which is broad enough to match ordinary conversation and unintentionally invoke the skill. In an agent that can modify configuration files, accidental activation increases the chance of unintended personality-test flows and, if confirmations are weak elsewhere, could contribute to unwanted config changes.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
77% confidence
Finding

The skill advertises file-modifying behavior such as backing up and modifying SOUL.md, yet the manifest does not declare any tool scope or permissions. Missing explicit scope makes the skill's effective capabilities opaque to users and reviewers, increasing the chance of unexpected environment or filesystem access if the underlying implementation uses env-capable code.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The natural-language description is entirely in Chinese and presents the skill as operating in that language, but it does not indicate that language selection is optional or that the skill is intentionally limited to a Chinese-speaking context. Under the policy, forcing a specific language without user opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The trigger phrase "人格测试" is generic and could match many unrelated user requests, causing the skill to activate unexpectedly. Over-broad activation can route conversations into the wrong skill context, creating confusion and potentially invoking capabilities the user did not intend to use.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

This markdown file uses Chinese throughout for user-facing instructions, triggers, acceptance criteria, and confirmation text. Under the policy rule, forcing a specific language without user opt-in is a natural-language policy concern unless the locale restriction is explicitly documented and justified, which is not present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This JavaScript file embeds natural-language agent configuration labels and style descriptions entirely in Chinese, starting with the header and continuing through generated SOUL.md content. Because the file provides no opt-in, fallback, or documented region-specific justification for this locale constraint, it may violate the language/locale policy for skills that should not force a specific language by default.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The rollback function restores a backup by reading a prior SOUL.md and overwriting the current file via writeSoul(), which is an irreversible state-changing operation for the active workspace file. Although the file header mentions backup and rollback functionality, the actual rollback path provides no confirmation prompt before replacing the current file contents.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The natural-language strings throughout the skill instruct and interact with the user only in Chinese, including prompts, confirmations, and error messages, with no opt-in or alternative language handling. This creates a language-policy concern because the skill effectively enforces a specific locale rather than offering a user choice or documenting a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The rollback path restores prior configuration immediately when invoked, with no in-flow confirmation or preview of what will change. In a conversational agent, ambiguous or accidental user input could trigger a state-changing restore operation, causing unintended reversion of configuration and potential loss of recent legitimate settings.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

A language/locale policy issue can arise when a skill or its user-facing instructions force a single language without offering choice or documenting a justified locale restriction. This document presents all headings, workflow text, and user interaction examples only in Chinese, with no indication that users may choose another language.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

This JSON file defines all user-facing dimension names and descriptions exclusively in Chinese, with no indication of locale choice, opt-in, or region-specific justification. Under the policy criteria, hard-coding a single language for user-facing text can be a natural-language locale violation when no user choice is offered.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

This JSON content presents all user-facing questions and answer text exclusively in Chinese. Under the policy for natural-language violations, forcing a specific language without offering a user choice or documenting a justified locale constraint can be a policy issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

This file contains user-facing natural-language strings only in Chinese across multiple result entries. Because the content appears fixed with no indication of user opt-in or language selection, it may violate the policy against forcing a specific language or locale.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

Natural-language strings in the file, including the title, comments, status messages, and CLI usage text, are presented only in Chinese. This imposes a specific language on users without any opt-in, alternative locale, or documented region-specific justification.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
85% confidence
Finding

The manifest describes taking an SBTI test, generating a personality report, selecting an Agent type, and configuring Agent SOUL.md. In addition to that workflow, this file exposes a restoration capability via handleRollback() that reverts to a previous version, which is broader than the described one-way configuration flow.

Content

No source excerpt is available for this finding.