Back to skill

Security audit

Home Buying

Security checks for vulnerabilities and agentic risk

Overview

This home-buying skill is a disclosed local workflow helper; it stores scoped notes locally and does not show hidden automation, external sharing, or transaction authority.

Before installing, decide whether you want persistent local home-buying memory at ~/home-buying/. Use explicit activation or one-transaction mode if you do not want this skill shaping every home-buying discussion, and avoid storing full identity documents, account numbers, or lender portal credentials.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · memory-template.md (reported line 15)May include surrounding context.

md
integration: pending

## Buy Box
<!-- Non-negotiables: location, property type, size, timeline -->
<!-- Deal-breakers and acceptable compromises -->

## Budget Guardrails

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · memory-template.md (reported line 20)May include surrounding context.

md
## Budget Guardrails
<!-- Monthly all-in ceiling and preferred range -->
<!-- Cash to close range and reserve target -->

## Financing Context
<!-- Pre-approval status, rate assumptions, lender constraints -->

Autonomous Decision Making

Medium
Category
Excessive Agency
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.

Content

Scanner excerpt · closing-readiness.md (reported line 21)May include surrounding context.

md
- Cash-to-close changed materially without explanation
- New title exception with unresolved risk
- Insurance premium shock that breaks affordability
- Repairs marked complete without verification

## Final Gate Decision

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The activation criteria are broad enough to trigger on general discussion of home buying, mortgages, or closing, which can cause the skill to engage and shape advice without clear user intent. In a domain involving major financial decisions and context collection, over-activation increases the risk of unwanted profiling, unnecessary persistence of sensitive preferences, and misleading automation in ordinary conversation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill directs the agent to persist home-buying context to a local memory file, but it does not require a clear user-facing notice and consent flow before storing potentially sensitive financial and housing preference data. Even though it says not to save secrets or unnecessary identifiers, the retained information can still reveal budget, geography, financing posture, and risk tolerance, creating privacy and confidentiality concerns.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.