ClawHub
Back to skill

Security audit

onehome

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent OneHome real-estate portal helper, but users should treat OneHome magic links and tokens as private credentials.

Install only if you intend to let the agent access your private OneHome portal. Treat full OneHome URLs, token parameters, bearer JWTs, and captured signed-in browser sessions as secrets; avoid pasting or sharing them where they may be logged, and prefer redacting token values in transcripts or outputs.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill explicitly instructs use of private OneHome magic-link URLs and bearer tokens, but it provides no warning that these credentials are sensitive and should not be exposed, logged, or shared. In this context, those tokens likely grant access to a buyer’s private portal data, so missing handling guidance materially increases the risk of credential leakage and unauthorized access.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
Four tools take `view: "compact" | "full"` — `onehome_get_by_address`,
`onehome_compare_properties`, `onehome_get_user` and `onehome_graphql` — and
**`compact` is the default**. You get the slim shape without asking for it;
an efficiency a caller has to request is one that is usually not requested.

**What compact does here is strip image and avatar URLs, and nothing else.**
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.

Static analysis

No suspicious patterns detected.