Back to skill

Security audit

fortune-master-suanming(算命)

Security checks for vulnerabilities and agentic risk

Overview

The skill is a mostly disclosed fortune-analysis skill, but it should be reviewed because it expands beyond its advertised BaZi scope and can handle sensitive personal data with under-scoped web lookup and high-stakes prediction guidance.

Install only if you are comfortable using a Chinese-language fortune-telling skill that asks for birth time, location, sex, relationship context, and possibly household details. Treat outputs as entertainment or cultural interpretation, not medical, legal, financial, investment, safety, or life-planning advice. Avoid sharing third-party personal data without consent, and do not allow web searches with identifying case details unless you explicitly approve them.

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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (55)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The guide materially expands the skill from 八字/四柱 analysis into tarot, divination hexagrams, synastry, and feng shui, which exceeds the declared scope in the manifest. Scope drift is dangerous because it can cause the agent to process additional categories of personal or sensitive data and perform actions users did not reasonably expect from the advertised skill.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The routing instructions explicitly broaden the skill from its declared BaZi/Four Pillars specialization into many unrelated divination systems such as Tarot, Ziwei Doushu, Western astrology, feng shui, and I Ching methods. This creates a scope-integrity failure: an orchestrator or user relying on the manifest may invoke the skill for domains it was not reviewed, constrained, or safety-checked to handle, increasing the chance of misleading outputs, policy bypass through undeclared capabilities, and incorrect downstream routing.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The changelog entry is written as Chinese-only content ('专注算命,八字学。') with no indication that the skill supports other languages or that this locale restriction is intentional and justified. The policy allows locale constraints only when users are given a choice or the restriction is clearly documented for a specific regional purpose.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

This markdown file presents the skill’s guidance, invocation example, and operating instructions only in Chinese. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale restriction is explicitly documented and justified.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill explicitly offers concrete year-by-year predictions for health, legal disputes, finances, and lifespan-related themes, which can encourage users to rely on unverifiable fortune-telling output for consequential real-world decisions. Although the file includes some boundary language and an entertainment disclaimer, those safeguards are undermined by the detailed predictive framing and are not attached closely enough to the risky sections to reliably prevent harmful reliance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The description on L04 is entirely in Chinese and presents the skill in that language by default. For a general skill manifest, this can constitute a language/locale policy issue because it does not indicate that users may choose another language or that the skill is intended only for Chinese-speaking users.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The file provides concrete financial and investment guidance, including expected income changes, timing for investing, and recommendations such as increasing investment or avoiding investment, without warning that this is not professional financial advice. Because the skill frames these predictions as detailed year-by-year guidance, users could make real financial decisions based on unverifiable fortune-telling output and suffer monetary loss.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file gives specific health predictions and quasi-medical recommendations such as likely symptoms, affected body systems, and advice on when to get checkups, but it does so without any disclaimer that the content is not medical advice. In a fortune-telling skill, users may over-trust these claims and delay appropriate diagnosis or treatment, especially when concrete years, symptoms, and body parts are presented as authoritative guidance.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The file provides concrete health-status, disease, and injury predictions such as specific organ risks, surgery, bloodshed, and relapse scenarios without any disclaimer that this is not medical advice and should not guide diagnosis or treatment. In a consumer-facing fortune-telling skill, users may over-trust these statements, delay appropriate care, or experience avoidable anxiety and harmful decision-making.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill predicts lawsuit exposure, imprisonment risk, and accident risk using authoritative language, but does not warn users that this is not legal advice and cannot assess real legal jeopardy. Users could make poor choices about compliance, disputes, reporting, or representation based on divination output rather than licensed legal counsel.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The guide encourages collection and analysis of highly sensitive personal data, including complete birth information for both parties and housing details, without any privacy warning, minimization rule, or handling guidance. In a fortune-analysis context, these details can directly identify individuals and reveal intimate relationship and household information, making misuse or overcollection particularly risky.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The document instructs network-based validation and research even though the skill is described as a Bazi analysis skill, not a web-researching or externally validating one. This creates undisclosed behavior expansion, increasing the risk of external data transmission, unbounded content retrieval, and analysis based on untrusted sources.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The guide directs network lookups based on user-provided cases but does not warn that details from those cases could be transmitted to external services. This is dangerous because users may share birth data, relationship data, or other sensitive case facts under the assumption the analysis remains local to the skill.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The guide operationalizes web research for fortune-analysis workflows without a clear necessity tied to the declared purpose. In this context, browsing is not a harmless enhancement: it can expose user case details to third parties and introduces untrusted external content into a skill handling personal life analysis.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file explicitly guides the skill to perform fate, fortune, and year-by-year predictive analysis, but only says to avoid medical topics and to provide humane guidance. It does not clearly warn users that such analysis is unverifiable, entertainment-like, culturally interpretive, and potentially psychologically influential, which can make vulnerable users over-rely on deterministic predictions about life decisions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This section instructs the skill to perform detailed astrological and personal-life analysis, including use of sensitive birth data, but does not require a disclaimer that the output is speculative or caution around privacy. That creates a real safety risk because users may over-trust non-factual guidance on relationships, career, finances, or health-adjacent life decisions, and may provide highly sensitive personal data without informed consent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The entire skill document is written in Chinese and does not indicate that users may choose another language or locale. Under the policy provided, a skill that effectively requires a specific language without user opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest describes a skill focused on 八字/四柱命理 analysis, including specific Bazi-related techniques and reasoning. This file materially broadens the documented analytical framework to additional metaphysical systems that are distinct from Bazi, indicating behavior/documentation scope beyond the stated skill description.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The manifest promises professional Bazi analysis centered on pillars, ten gods, useful god selection, combinations/clashes, and luck cycles. The explicit workflow here instructs using Feng Shui for layout advice and Qimen/Liuren for timing/date selection, which are separate practices rather than obvious implementation details of Bazi analysis.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This markdown file contains only Chinese-language instructions and examples, with no indication that the user can choose another language or that the skill is intentionally limited to Chinese-speaking users. 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.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The document promises '综合判断 → 具体应事(精确到年月)' and later a report structure with precise life-event judgment, but includes no caution about uncertainty, emotional impact, or the risk of users over-trusting deterministic predictions. In a fortune-analysis skill, this raises meaningful safety concerns because users may make relationship, financial, health, or family decisions based on authoritative-sounding forecasts.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The framework explicitly processes highly sensitive personal data such as birth time, birthplace-derived true solar time, family relationships, health, and predicted life events, but provides no privacy notice, consent guidance, retention limits, or caution about handling sensitive inferences. In this skill context, that is a real privacy and safety issue because the system is designed to derive intimate personal conclusions from user-supplied data and present them as authoritative analysis.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

This file gives highly specific predictive guidance about health, finances, relationships, and major life decisions, and frames outcomes as actionable and foreseeable without any uncertainty disclaimer or warning not to rely on it for professional advice. In the context of a fortune-telling skill, that increases risk because users may treat the content as authoritative guidance for medical, financial, or consequential personal decisions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This markdown file presents all instructional content in Chinese and does not indicate that the user can choose another language. Under the language/locale policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale limitation is explicitly documented and justified.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The invocation guidance is broad enough that ordinary user requests like '看看今年的运势' could activate the skill implicitly, even when the user did not clearly request this specific skill. In agent environments with multiple installed skills, overly permissive triggers can cause unintended routing, leading to surprise collection of sensitive personal data such as birth date, time, sex, and location.

Content

No source excerpt is available for this finding.