Security audit

jf-open-pro-ai-child-care

Security checks for vulnerabilities and agentic risk

Overview

The skill is coherent for JFTech child-care monitoring, but it needs review because it handles child/home surveillance data and can change monitoring or person-library state without prominent privacy or confirmation safeguards.

Review before installing. Use this only if you trust the JFTech API integration, are authorized to administer the device and child-care service, and understand that credentials, device identifiers, alarm data, behavior statistics, and person-library changes may be sent to JFTech endpoints. Treat the app secret and authorization token as sensitive, and manually confirm any enable/disable or stranger remove action before running it.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (8)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 93% confidence
Finding: The skill documentation instructs users to run Python scripts that perform outbound API calls and shell commands, yet no explicit permissions model is declared. This creates a transparency and consent gap: a user or host platform may invoke a skill with network and shell capabilities without clear prior disclosure, increasing the risk of credential misuse and unintended external communications.

Context-Inappropriate Capability

Medium

Confidence: 84% confidence
Finding: The code performs authenticated remote modification of a biometric/person repository through add/remove APIs, but this capability is not clearly justified by the stated monitoring-only purpose. In a child-care context, unauthorized or unexpected manipulation of stranger records could alter alerts, surveillance outcomes, or privacy-sensitive person data without clear user awareness.

Vague Triggers

Medium

Confidence: 83% confidence
Finding: The trigger phrases include broad everyday terms such as '异常告警' and '行为统计', which can cause accidental invocation outside a clearly intended context. For a skill that accesses surveillance, alarm, and child-monitoring data, unintended activation increases privacy exposure and the chance of state-changing actions being initiated without deliberate user intent.

Missing User Warnings

High

Confidence: 97% confidence
Finding: This skill is explicitly designed for continuous household monitoring of children, including behavior statistics and alarm processing, but the documentation lacks a clear privacy notice, retention/use disclosure, or impact warning. Because the data concerns minors and in-home surveillance, the absence of explicit informed-consent language materially increases privacy, safety, and compliance risk.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The documented capabilities include state-changing operations such as enabling/disabling child-care monitoring and adding or removing entries from a stranger/person library, but there is no prominent warning about the consequences of those actions. In this context, silent or insufficiently signposted state changes can degrade monitoring coverage, alter alerting behavior, or affect how people are classified in a home-surveillance workflow.

Missing User Warnings

Medium

Confidence: 84% confidence
Finding: The script performs a remote state-changing operation that enables or disables a child-care monitoring feature immediately when invoked, with no confirmation prompt, dry-run mode, or explicit warning to the operator. In a security-sensitive context involving child monitoring and device/account settings, this increases the risk of accidental or unauthorized changes if the command is triggered by mistake, by automation, or with misrouted credentials.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The script requires highly sensitive values such as appKey, secret, UUID, Authorization token, and device identifiers, then transmits them to a remote third-party API at execution time without any runtime disclosure, confirmation prompt, or safety notice. In this childcare-monitoring context, the data relates to a child surveillance service and household device activity, so silent transmission increases privacy and credential-exposure risk if the script is run by an unsuspecting operator or in shared environments where CLI arguments may be logged.

Missing User Warnings

Medium

Confidence: 81% confidence
Finding: The remove operation triggers a destructive remote action immediately from command-line input with no confirmation, dry-run, or secondary validation. In a system managing person/stranger records for child safety monitoring, accidental deletion could suppress recognition state or disrupt safety workflows and auditability.

Static analysis

No suspicious patterns detected.