Back to skill

Security audit

奇门遁甲排盘技能

Security checks for vulnerabilities and agentic risk

Overview

This is a local Chinese divination charting skill with some overstated accuracy claims and an undeclared backup utility, but no evidence of hidden execution, exfiltration, persistence, or credential access.

Install only if you want a Chinese-language local 奇门遁甲 charting tool. Treat the outputs as cultural or entertainment guidance, not professional medical, legal, financial, or safety advice. Avoid running scripts/backup.sh unless you understand that it writes under your OpenClaw skills directory and prunes older backup_* folders.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (19)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

声明描述的是一个命理排盘/解盘能力,但提供的代码并未实现任何排盘、起局、八门九星八神计算或解盘分析逻辑。相反,它只进行本地文件备份、生成备份信息并清理旧备份,属于维护/运维脚本。该行为与声明的主要用途明显不同,且涉及文件复制与删除等未声明的能力,因此应判定为描述与实际行为不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

该代码的主要用途与声明的大方向一致,确实是在做奇门遁甲排盘,并包含阴阳遁、局数、值符值使、九宫门星神落位等核心输出,也没有发现未声明的敏感能力、外部资源访问或异常触发行为。但声明中提到“断卦解盘自动化分析”,而代码实际上只返回排盘结果和几条固定的“用神参考”,没有针对盘面进行规则推断、吉凶判断、事项分析或自动解读,因此这一能力存在明显描述超出。另一个次要问题是代码多处标注为‘简化算法’‘近似值’,说明它更像基础排盘/演示工具,而非完整严谨的专业排盘分析器。综合判断为描述与实际行为存在不小偏差,应标记为 mismatch。

Content

No source excerpt is available for this finding.

Chaining Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

The pipeline ls -1dt backup_* | tail ... | xargs rm -rf chains filename discovery directly into recursive deletion. This pattern is dangerous because unusual filenames, whitespace/newlines, or unexpected matches can cause incorrect arguments to be passed to rm -rf, increasing the risk of deleting unintended paths.

Content

Scanner excerpt · scripts/backup.sh (reported line 61)May include surrounding context.

sh
if [ "$BACKUP_COUNT" -gt "$MAX_BACKUPS" ]; then
        echo "清理旧备份(保留最近 $MAX_BACKUPS 个)..."
        ls -1dt backup_* | tail -n +$((MAX_BACKUPS + 1)) | xargs rm -rf
        echo "清理完成"
    fi
else

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The JieQi calculator explicitly promises ±1 minute precision while using a fixed approximate date table and comments acknowledging it is simplified. In this skill, solar-term boundaries drive yin/yang dun and bureau selection, so incorrect results can propagate through the entire chart and automated interpretation, making the false precision claim materially harmful.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

This markdown file uses Chinese throughout, including headings, usage guidance, and output examples, with no indication that users may choose another language or that the skill is intended only for a Chinese-language or region-specific audience. Under the policy for natural-language violations, forcing a specific language without user opt-in is reportable.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The entire skill description, usage guidance, CLI help examples, and output schema are presented exclusively in Chinese, with no indication that users may choose another language or that the skill is intentionally limited to a Chinese-speaking context. Under the policy, a language/locale constraint should be opt-in or clearly justified; neither is present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file presents the skill name, description, features, and usage entirely in Chinese, which can force a specific language experience on users without opt-in. Under the policy, language restrictions should either offer user choice or be explicitly justified as region- or locale-specific.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest describes a 奇门遁甲排盘 skill that computes chart layouts and automated interpretation, but this script creates backup directories, copies skill files, writes backup metadata, and deletes older backups. Those maintenance operations are not part of chart generation or analysis and materially extend the skill's behavior beyond its declared purpose.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The script includes deletion of older backup directories using a retention policy, which is destructive behavior. Although the deletion is intended to manage backups, it is still potentially dangerous because it removes data automatically and is not essential to the stated divination/charting functionality of the skill.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Old backup directories are deleted with rm -rf without any confirmation, dry-run, or explicit warning to the user. If the path set is broader than expected or naming collisions occur, this can cause irreversible data loss with little opportunity for the user to intervene.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest description says the skill supports '断卦解盘自动化分析', which implies substantive automated reading or interpretation of the chart. In code, qimen_pan only assembles calculated fields and format_output appends generic static '用神参考' lines; there is no chart-specific interpretive analysis logic.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The module-level documentation claims '高精度' and exact astronomical/lunar capabilities, but multiple implementations later use admitted simplifications and heuristics. This is dangerous because downstream agents or users may treat the output as authoritative for time-sensitive or decision-support workflows, creating integrity and trust risks through misleading claims rather than code execution.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

This Python file is a code file, so only SQP-2 and SQP-3 apply. The docstrings, CLI help text, and formatted user-facing output are all fixed in Chinese, and the skill does not provide any opt-in, configuration, or justification that it is intentionally limited to a Chinese-only audience or region-specific compliance context.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The file provides detailed divination guidance and interpretive rules without any disclaimer about uncertainty, non-scientific reliability, or limits on use for important life decisions. In the context of an automation-oriented skill that supports analysis and interpretation, users may over-trust outputs for health, legal, financial, or other consequential decisions, increasing risk of harm from misleading guidance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The natural-language content forces a single language/locale experience by presenting all instructions and reference material only in Chinese. Under SQP-3, this is a policy concern unless the skill explicitly offers language choice or documents that the locale restriction is intentional and justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

Comments and console output are written entirely in Chinese, including operational status and failure messages. This imposes a language choice on users without opt-in or explanation, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The file’s natural-language description and all user-facing CLI text are written only in Chinese, and the script provides no option to select another language or opt into this locale. Under the policy for natural-language violations, a fixed language without user choice can be a locale policy issue unless it is clearly justified as region-specific.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
98% confidence
Finding

The docstring at get_jieqi states it returns '(当前节气,下一节气,是否阳遁)', which is a 3-tuple including the next solar term. The implementation returns only 'jieqi, is_yang_dun', and callers also unpack only two values, so the documentation actively contradicts the function behavior.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The module header describes 'v3.0.0' and adds v3.0.0 analysis capabilities, but the main function docstring and CLI parser description still label the tool as 'v2.0.1'. This creates contradictory developer-facing documentation about what version and feature set the code represents.

Content

No source excerpt is available for this finding.