Back to skill

Security audit

hyperbolic-discounting

Security checks for vulnerabilities and agentic risk

Overview

The skill gives useful coaching but directs financial penalties, access restrictions and public accountability without clear approval boundaries.

Use this skill for planning and coaching. Require your approval before any payment, enrollment, access restriction or disclosure, including stronger replacements. Agree on costs, recipients, duration and cancellation before implementation.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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
Findings (1)

The workflow instructs installation of binding commitment devices without warning about financial, privacy, or access consequences.

Medium
Category
Output Handling
Confidence
92% confidence
Finding

The workflow moves from recommending commitment devices to installing financial penalties, restrictive defaults, and public accountability without explicitly requiring informed authorization before consequential actions. A tool-enabled agent could interpret this as permission to incur charges, restrict access, or disclose personal goals; the requirement that a device be acceptable to the user reduces concern but does not establish consent to specific actions and consequences. This is an unsafe execution-boundary omission, not evidence of malicious code or exfiltration.

Content

No source excerpt is available for this finding.