Security audit

Emergence Superorganism Novel

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent creative-writing skill with no executable code, persistence, credential access, or hidden data flow.

Install this if you want strong, opinionated guidance for superorganism fiction. Review the stylistic rules first, especially the dehumanizing terminology and Chinese examples, and explicitly tell your agent when you want a different language, tone, or treatment of human characters.

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
Rogue AgentSelf-Modification, Session Persistence
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (6)

Self-Modification

High

Category: Rogue Agent
Confidence: 90% confidence
Finding: Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.
Content: Scanner excerpt · SKILL.md (reported line 248)May include surrounding context.
md
No superorganism is immortal in practice. Decline sets in when the coordination rules that enabled growth become maladaptive: - **Rule rigidity**: The superorganism cannot change its own rules fast enough. (Legalism in the Qin Empire — "失期当斩" applied without exception until everyone was in revolt.) - **Internal contradiction**: The parts that make the superorganism strong begin to work against each other. (A company whose sales team and engineering team optimize for conflicting metrics.) - **Mirror loss**: The peer entities that defined the superorganism's identity disappear or transform, leaving the superorganism without a clear sense of self. (A nation after its historical rival collapses.) - **Node alienation**: The human nodes stop identifying with the superorganism. They obey without belief. Then they stop obeying.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: 文件名和全文内容表明该技能说明固定为中文，但文档中没有提供语言选择、用户可切换语言的说明，或明确的地区/受众限定。根据规则，强制特定语言且未获用户选择同意，属于自然语言层面的语言/locale 政策问题。
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The skill explicitly instructs the agent to refer to humans using prescribed Chinese terms such as 'ports, nodes, cells' and to use specific wording like 它 for humans in some contexts. That overrides normal user-controlled style and language choice, and can cause unwanted output behavior, especially for users who did not opt into Chinese or dehumanizing terminology. In a creative-writing skill this is less dangerous than code execution, but it is still a policy and output-shaping issue because it constrains language in a way that may conflict with user intent and safety expectations.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The template's only opening-line examples are in Chinese, which can steer an agent to produce Chinese output even when the user did not request that language. This is a quality and instruction-alignment issue rather than a direct security exploit, but it can cause unwanted language switching, degraded usability, and failure to honor user preferences.
Content: No source excerpt is available for this finding.

YARA rule 'network_reconnaissance': Network reconnaissance and scanning patterns [hacktools]

Medium

Category: YARA Match
Confidence: 65% confidence
Finding: YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).
Content: Scanner excerpt · templates/rome-legion-memoir-part1.md (reported line 225)May include surrounding context.
md
rtbeats. The Thirteenth, relieved, spoke its first words: "I thought you'd be slower." "I thought you'd be quieter," I said. "Impossible," the Thirteenth said. "I stopped being quiet the day the Fourteenth died." That evening Vercingetorix rode down the hill to surrender. He stripped off his armor and sat on the ground. I looked down at him — he was the will of an entire people, a powerful, fierce, but brief light. I thought of the Fourteenth. The Sabis. Gergovia. "This enemy," the Eighth said, standing beside me. "Do you respect him." "I do," I said. "But not because he is Vercingetorix. Because he made **me** into what I am today." Yes. The Gallic War was over. The one called Caesar had won. But the one who won was not him. It was me — I had learned to lose, learned to save, learned to dismantle my own bridge and rebuild myself from the same timbers. I did not yet know what would happen next. The Rubicon was still waiting for me in the south. Pompey's legions — my own o

Vague Triggers

Low

Category: Not specified by scanner
Confidence: 83% confidence
Finding: The usage section says to invoke the skill when the user wants fiction from the perspective of several broad categories, culminating in 'any distributed system' with a conceptual condition. This leaves the trigger scope open-ended and does not provide exclusion cases or stricter phrasing, which could cause unintended invocation for generic writing requests about cities, companies, or ideologies.
Content: No source excerpt is available for this finding.