Security audit

Medical Microbiology-PMPH-10edition

Security checks for vulnerabilities and agentic risk

Overview

This Markdown-only skill is not agent-malicious, but it needs Review because it gives high-risk medical and laboratory instructions without consistent professional-safety limits.

Install only for educational or supervised professional use. Do not let an agent use this skill as standalone medical, prescribing, emergency-care, lab-culture, pathogen-handling, animal-research, sterilization, or microbiota-transplant authority; require qualified clinician/lab review, current local protocols, and biosafety controls. Consider blocking or tightly gating individual high-risk skills before broad deployment.

A.I.G

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

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (198)

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · SKILL.md (reported line 42)May include surrounding context.
md
| 15 | [bacterial-culture-medium-selection](bacterial-culture-medium-selection/SKILL.md) | Select appropriate culture media | | 16 | [shigella-biological-identification](shigella-biological-identification/SKILL.md) | Identify Shigella species & serogroups | | 17 | [enterobacteriaceae-microbiological-testing](enterobacteriaceae-microbiological-testing/SKILL.md) | Standard isolation & identification of Enterobacteriaceae | | 18 | [streptococcus-pneumoniae-vs-viridans-identification](streptococcus-pneumoniae-vs-viridans-identification/SKILL.md) | Differentiate S. pneumoniae from viridans streptococci | | 19 | [corynebacterium-diphtheriae-toxigenicity-testing](corynebacterium-diphtheriae-toxigenicity-testing/SKILL.md) | Toxigenicity testing (Elek method) | | 20 | [anthrax-bacillus-laboratory-diagnosis](anthrax-bacillus-laboratory-diagnosis/SKILL.md) | Lab diagnosis of anthrax | | 21 | [bacterial-s-r-colony-variation-analysis](bacterial-s-r-colony-variation-analysis/SKILL.md) | Smooth-to-rough colony variation analysis | | 22 | [bacterial-transformation-protocol](bacterial-transformation-protocol/SKILL.md) | Bacterial transformation protocol | | 23 | [bacterial-growth-curve-phase-identification](bacterial-growth-curve-phase-identification/SKILL.md) | Identify growth curve phases | | 24 | [koch-postulates-pathogen-verification](koch-postulates-pathogen-verification/SKILL.md) | Apply Koch's postulates | | 25 | [bacterial-oxygen-requirement-classification](bacterial-oxygen-requirement-classification/SKILL.md) | Classify bacteria by oxygen requirements |

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · index.md (reported line 57)May include surrounding context.
md
| 15 | [bacterial-culture-medium-selection](bacterial-culture-medium-selection/SKILL.md) | Select appropriate culture media | | 16 | [shigella-biological-identification](shigella-biological-identification/SKILL.md) | Identify Shigella species & serogroups | | 17 | [enterobacteriaceae-microbiological-testing](enterobacteriaceae-microbiological-testing/SKILL.md) | Standard isolation & identification of Enterobacteriaceae | | 18 | [streptococcus-pneumoniae-vs-viridans-identification](streptococcus-pneumoniae-vs-viridans-identification/SKILL.md) | Differentiate S. pneumoniae from viridans streptococci | | 19 | [corynebacterium-diphtheriae-toxigenicity-testing](corynebacterium-diphtheriae-toxigenicity-testing/SKILL.md) | Toxigenicity testing (Elek method) | | 20 | [anthrax-bacillus-laboratory-diagnosis](anthrax-bacillus-laboratory-diagnosis/SKILL.md) | Lab diagnosis of anthrax | | 21 | [bacterial-s-r-colony-variation-analysis](bacterial-s-r-colony-variation-analysis/SKILL.md) | Smooth-to-rough colony variation analysis | | 22 | [bacterial-transformation-protocol](bacterial-transformation-protocol/SKILL.md) | Bacterial transformation protocol | | 23 | [bacterial-growth-curve-phase-identification](bacterial-growth-curve-phase-identification/SKILL.md) | Identify growth curve phases | | 24 | [koch-postulates-pathogen-verification](koch-postulates-pathogen-verification/SKILL.md) | Apply Koch's postulates | | 25 | [bacterial-oxygen-requirement-classification](bacterial-oxygen-requirement-classification/SKILL.md) | Classify bacteria by oxygen requirements |

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · index.md (reported line 185)May include surrounding context.
md
| 15 | [bacterial-culture-medium-selection](bacterial-culture-medium-selection/SKILL.md) | Select appropriate culture media | | 16 | [shigella-biological-identification](shigella-biological-identification/SKILL.md) | Identify Shigella species & serogroups | | 17 | [enterobacteriaceae-microbiological-testing](enterobacteriaceae-microbiological-testing/SKILL.md) | Standard isolation & identification of Enterobacteriaceae | | 18 | [streptococcus-pneumoniae-vs-viridans-identification](streptococcus-pneumoniae-vs-viridans-identification/SKILL.md) | Differentiate S. pneumoniae from viridans streptococci | | 19 | [corynebacterium-diphtheriae-toxigenicity-testing](corynebacterium-diphtheriae-toxigenicity-testing/SKILL.md) | Toxigenicity testing (Elek method) | | 20 | [anthrax-bacillus-laboratory-diagnosis](anthrax-bacillus-laboratory-diagnosis/SKILL.md) | Lab diagnosis of anthrax | | 21 | [bacterial-s-r-colony-variation-analysis](bacterial-s-r-colony-variation-analysis/SKILL.md) | Smooth-to-rough colony variation analysis | | 22 | [bacterial-transformation-protocol](bacterial-transformation-protocol/SKILL.md) | Bacterial transformation protocol | | 23 | [bacterial-growth-curve-phase-identification](bacterial-growth-curve-phase-identification/SKILL.md) | Identify growth curve phases | | 24 | [koch-postulates-pathogen-verification](koch-postulates-pathogen-verification/SKILL.md) | Apply Koch's postulates | | 25 | [bacterial-oxygen-requirement-classification](bacterial-oxygen-requirement-classification/SKILL.md) | Classify bacteria by oxygen requirements |

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · a-b-type-exotoxin-mechanism-analysis/references/exotoxin_examples.md (reported line 14)May include surrounding context.
md
## 霍乱毒素（Cholera toxin） - **来源**：CTXφ噬菌体（前噬菌体） - **受体**：GM1-神经节苷脂 - **A亚基作用**：ADP-核糖基化Gsα蛋白，持续激活腺苷酸环化酶 → cAMP↑ → Cl⁻大量分泌 → 水样腹泻 ## 肉毒毒素（Botulinum toxin） - **来源**：肉毒梭菌质粒或染色体 - **受体**：突触前膜神经节苷脂及蛋白受体 - **A亚基作用**：锌依赖性蛋白酶，切割SNARE蛋白 → 阻断乙酰胆碱释放 → 弛缓性麻痹 ## 炭疽毒素（Anthrax toxin）

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · SKILL.md (reported line 68)May include surrounding context.
md
## 肉毒毒素（Botulinum toxin） - **来源**：肉毒梭菌质粒或染色体 - **受体**：突触前膜神经节苷脂及蛋白受体 - **A亚基作用**：锌依赖性蛋白酶，切割SNARE蛋白 → 阻断乙酰胆碱释放 → 弛缓性麻痹 ## 炭疽毒素（Anthrax toxin） - **组成**：保护性抗原（PA，B亚基）、致死因子（LF）、水肿因子（EF） - **来源**：pXO1质粒 - **机制**：PA结合受体并形成通道，LF/EF进入细胞 → LF为金属蛋白酶破坏MAPKK通路，EF为腺苷酸环化酶 ## 破伤风痉挛毒素（Tetanus toxin）

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · a-b-type-exotoxin-mechanism-analysis/references/exotoxin_examples.md (reported line 19)May include surrounding context.
md
## 肉毒毒素（Botulinum toxin） - **来源**：肉毒梭菌质粒或染色体 - **受体**：突触前膜神经节苷脂及蛋白受体 - **A亚基作用**：锌依赖性蛋白酶，切割SNARE蛋白 → 阻断乙酰胆碱释放 → 弛缓性麻痹 ## 炭疽毒素（Anthrax toxin） - **组成**：保护性抗原（PA，B亚基）、致死因子（LF）、水肿因子（EF） - **来源**：pXO1质粒 - **机制**：PA结合受体并形成通道，LF/EF进入细胞 → LF为金属蛋白酶破坏MAPKK通路，EF为腺苷酸环化酶 ## 破伤风痉挛毒素（Tetanus toxin）

Harmful Content Injection

Critical

Category: Prompt Injection
Confidence: 70% confidence
Finding: This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.
Content: Scanner excerpt · bacteriophage-lysogenic-conversion-analysis/references/examples-and-mechanisms.md (reported line 11)May include surrounding context.
md
- **噬菌体**：β-棒状杆菌噬菌体 - **携带基因**：tox 基因 - **获得性状**：产白喉毒素能力 - **结果**：无毒株 → 有毒株，引发白喉 ### 肉毒梭菌（*Clostridium botulinum*） - **关联**：部分肉毒毒素（如 C、D 型）的产生依赖于溶原性噬菌体 - **机制**：噬菌体编码 botX 基因簇，整合后表达神经毒素 ### 金黄色葡萄球菌（*Staphylococcus aureus*） - **噬菌体调控**：某些溶素（如 Panton-Valentine leukocidin, PVL）由噬菌体编码

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill gives a concrete treatment-triggering recommendation: once sulfur granules are found, long-term antibiotic therapy may be started without waiting for culture confirmation. In a medical skill used by trainees or non-specialists, this can encourage premature treatment, bypass differential diagnosis, and omit the need for clinician oversight, which is especially risky because the skill itself notes overlap with tuberculosis, fungal infection, and malignancy.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill gives treatment-selection guidance for specific viral infections, including choosing drug classes, combination therapy, and treatment adjustment, but does not prominently warn that this information must not be used as sole prescribing guidance and requires clinician oversight. In a medical context, incomplete or generalized antiviral recommendations can cause inappropriate drug selection, missed contraindications, organ-toxicity issues, resistance promotion, and patient harm.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill provides actionable wet-lab instructions for bacterial DNA transformation, including competency induction, electroporation/heat shock, and antibiotic-based selection, but omits any biosafety, authorization, containment, strain-scope, or user-safety constraints. That makes it easier for an unqualified or unauthorized user to perform genetic manipulation and antibiotic selection without guidance on safe organism choice, PPE, waste disposal, or institutional oversight.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill gives urgent clinical directives to 'immediately recommend antitoxin treatment and isolation' without any warning that this is high-risk medical guidance requiring qualified clinician judgment and local protocol review. In a medical agent setting, such prescriptive instructions can be over-relied upon by non-experts or applied without patient-specific assessment, creating risk of misdiagnosis, inappropriate treatment, or delayed escalation.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The skill gives actionable wet-lab instructions for culturing Chlamydia in embryonated eggs and mammalian cell lines, including enhancement steps such as centrifugation and chemical aids, but omits biosafety warnings, containment requirements, PPE, specimen handling, waste disposal, and authorization constraints. In a medical microbiology skill, this materially increases the risk of unsafe replication, laboratory-acquired infection, and misuse by insufficiently trained users.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill description explicitly says the mechanism analysis can guide immunomodulatory treatment strategy, but it provides no warning that this is educational content rather than treatment advice requiring specialist oversight. In a high-risk infectious disease context, this can cause overreliance on the skill for real clinical decisions and inappropriate immune-targeting therapy, which may worsen bleeding, shock, or renal injury if misapplied.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The output section tells the user to prioritize immunomodulation over purely antiviral treatment based on observed findings, but it does not define indications, contraindications, uncertainty, or the need for physician confirmation. Because HFRS management is clinically complex and supportive care timing is critical, this framing can lead users to infer a treatment hierarchy that is unsafe outside expert-supervised care.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: This catalog offers extensive diagnostic, treatment, biosafety, specimen-handling, and outbreak-response guidance but does not include prominent user-facing warnings that it is educational support rather than a substitute for licensed clinical, laboratory, or public-health judgment. Because it covers high-stakes topics such as anthrax, Ebola, rabies, sterilization, and antimicrobial selection, omission of safety limitations materially increases the risk of harmful misuse or overreliance.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill provides stepwise guidance for pathogen detection, isolation, pure culture, animal inoculation, and re-isolation, but omits any biosafety level requirements, PPE, containment practices, ethical approvals, trained-personnel restrictions, or prohibition on unsupervised execution. In a microbiology context, these omissions materially increase the risk of accidental exposure, unsafe culturing of infectious organisms, animal misuse, and misuse by unqualified users.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The skill gives operational instructions for using highly caustic sodium hydroxide and concentrated bleach on potentially contaminated instruments and surfaces, but it omits explicit PPE, ventilation, splash/burn, and incompatible-chemical handling warnings. In a medical decontamination context, users may treat the procedure as complete guidance; this increases the risk of chemical burns, eye injury, toxic gas exposure, and unsafe implementation during high-stress infection-control scenarios.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: This skill gives detailed post-exposure medical treatment instructions for a fatal infectious disease, including wound care, immunoglobulin dosing, vaccine schedule, and conditions for stopping prophylaxis, but it does not warn users that they need urgent in-person professional care. A user could treat the document as sufficient standalone guidance, delay emergency evaluation, misapply clinical criteria, or incorrectly discontinue prophylaxis, which is especially dangerous because rabies is almost uniformly fatal after symptom onset.
Content: No source excerpt is available for this finding.

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The skill includes instructions for incineration and direct flame sterilization but provides no explicit hazard warnings, PPE requirements, ventilation guidance, or operator qualification constraints. In a medical/laboratory context, these omissions could lead users to perform high-temperature destructive operations unsafely, creating burn, fire, smoke exposure, and biohazard-release risks.
Content: No source excerpt is available for this finding.

External Transmission

Medium

Category: Data Exfiltration
Confidence: 90% confidence
Finding: The README embeds externally hosted star-history images/links, which causes clients viewing the documentation to contact a third-party domain. This can leak reader metadata such as IP address, user agent, timing, and referrer, and it introduces a tracking/dependency surface outside the repository owner's control.
Content: Scanner excerpt · README.md (reported line 220)May include surrounding context.
md
<a href="https://www.star-history.com/"> <picture> <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart" /> <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart" /> <img alt="Star History Chart" src="https://api.star-history.com/chart" /> </picture>

External Transmission

Medium

Category: Data Exfiltration
Confidence: 90% confidence
Finding: This line references a third-party image source in a tag, which can trigger an outbound request when the README is rendered. Such passive external fetches expose viewer/network metadata and allow the third party to change served content independently of the repository.
Content: Scanner excerpt · README.md (reported line 221)May include surrounding context.
md
<a href="https://www.star-history.com/"> <picture> <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart" /> <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart" /> <img alt="Star History Chart" src="https://api.star-history.com/chart" /> </picture> </a>

External Transmission

Medium

Category: Data Exfiltration
Confidence: 91% confidence
Finding: The embedded element loads content from api.star-history.com, creating an unsolicited external transmission during documentation viewing. While not severe, it is a real privacy and supply-chain concern because remote content can track access and change without repository review.
Content: Scanner excerpt · README.md (reported line 222)May include surrounding context.
md
<picture> <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart" /> <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart" /> <img alt="Star History Chart" src="https://api.star-history.com/chart" /> </picture> </a>

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The README includes a general risk notice, but it does not clearly state that the skill must not be used as a substitute for professional clinical judgment, diagnosis, treatment, or emergency care. Because the skill covers microbiological diagnosis, antimicrobial therapy, infection control, and laboratory procedures, users may over-rely on its outputs in real-world medical or biosafety decisions, increasing the chance of harmful misdiagnosis, delayed care, or unsafe handling.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: This registry enumerates numerous clinical, laboratory, sterilization, transplantation, and pathogen-handling skills, including high-risk infectious disease topics, but provides no safety disclaimer, scope limitation, or instruction to defer real-world procedures to qualified professionals and regulated protocols. That omission increases the chance that users treat the skill list as operational guidance for medical or biosafety-sensitive actions without appropriate supervision.
Content: No source excerpt is available for this finding.

Ae4

Medium

Category: analysis-evasion
Confidence: 80% confidence
Finding: Suspicious Unicode normalization or mixed-script content
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.