Security audit

New York City

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly ordinary NYC guidance, but it has a real privacy concern because some files allow sensitive personal or financial details to be saved in plaintext despite the main file saying they are not stored.

Install only if you are comfortable with optional local memory under ~/new-york-city/. Do not ask the skill to save passwords, account numbers, passport details, payment information, or exact home addresses, and expect that some official NYC-related websites beyond the main endpoint table may be contacted for current guidance.

A.I.G

Vulnerability Patterns

Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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

Findings (2)

T09 · Insecure Skill Coding Practices

Warning

Location: setup.md:57
Finding: Plaintext Persistent Memory May Contain Highly Sensitive Information
Content: View full analysis
Remediation: View remediation

other

Note

Location: SKILL.md:74
Finding: External Endpoint Disclosure Omits Permitted Third-Party Domains
Content: View full analysis
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Autonomous Decision Making

Medium

Category: Excessive Agency
Confidence: 85% confidence
Finding: Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
Content: Scanner excerpt · moving-and-housing.md (reported line 29)May include surrounding context.
md
## Apartment Search Rules ### 1. Documentation and timing matter - Good listings move fast, but pressure is not a reason to skip verification. - Clarify income, guarantor, international-doc, roommate, and timing constraints first. ### 2. Building reality beats listing photos

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The activation rule is intentionally very broad: "whenever the user mentions New York City at all." This can cause the skill to trigger outside its appropriate scope, leading to irrelevant instruction injection into unrelated conversations and potentially unnecessary collection or persistence of user context tied to NYC topics.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.