Back to skill

Security audit

Couple

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent local relationship-planning skill; its main risk is that personal notes persist in the user’s home directory.

Install only if you are comfortable keeping relationship notes in local files under ~/couple/. Avoid storing private partner information, credentials, health, financial, intimate, or non-consensual details, and be careful on shared computers or systems with cloud backup enabled.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (4)

Hidden Instructions

High
Category
Prompt Injection
Content
- Timezone(s): [if relevant]

## Partner Preferences
<!-- Only what they've shared directly -->
- Loves:
- Dislikes:
- Gift style: [experiences/physical/words]
Confidence
70% confidence
Finding
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
| Health data | Conditions, cycles | Coercion risk |
| Intimate details | Sexual history | Weaponizable |
| Credentials | Passwords, PINs | Autonomy violation |
| Partner data without consent | Anything not shared by them | Surveillance |

## Requests to Refuse
Confidence
75% 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.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The template encourages storing sensitive relationship data such as anniversaries, birthdays, preferences, milestones, and personal history in a local file without any warning about privacy, retention, device security, or consent. Even though the storage is local, this can expose intimate personal information if the device, home directory, backups, or synced files are accessed by others.

Missing User Warnings

Low
Confidence
93% confidence
Finding
The skill stores relationship-related data under `~/couple/`, but the user-facing description does not clearly warn that personal data will be persisted in the home directory. Because this skill handles sensitive interpersonal context, lack of explicit disclosure can lead users to share more information than intended and increases privacy risk on shared or unmanaged systems.

Static analysis

No suspicious patterns detected.