Security audit

更适合中国宝宝体质的全程求职陪跑skill

Security checks across malware telemetry and agentic risk

Overview

JobHound is a coherent job-search assistant that saves sensitive career notes locally, so it should be used only in a private workspace.

Install only if you are comfortable with the agent storing resume details, application status, interview notes, weaknesses, offer information, and salary expectations in local workspace files. Use a private workspace, review generated files before sharing them, and delete MEMORY.md, output, report, and application-tracking files when you no longer want the data retained.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The activation phrases are very broad (e.g., generic job-seeking and interview-prep language), so the skill may trigger unintentionally during ordinary conversation. That can cause unsolicited processing of resumes, job-search context, or memory files, which increases privacy and integrity risk even without explicit malicious intent.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger section defines broad categories instead of a precise activation rule, making it ambiguous when the skill should run. Ambiguous invocation is dangerous because it can lead to unexpected file reads/writes and collection of sensitive career data without sufficiently clear user intent.

Missing User Warnings

High

Confidence: 98% confidence
Finding: The skill mandates cross-session persistence of detailed candidate profiles, application progress, interview weaknesses, and story banks in MEMORY.md without any visible consent, retention, or sensitivity controls. This is dangerous because it creates a durable store of personal and potentially confidential employment data that could be exposed, over-retained, or reused beyond the user's expectations.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The skill automatically generates and updates multiple local files containing personal job-search data, but does not tell the user that file writes will occur. Undisclosed persistence is risky because users may reveal resumes, target roles, salary expectations, or other sensitive information without understanding it will be saved locally.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill requires searching external platforms for interview experiences and salary information without notifying the user about network use or potential privacy implications. This can expose user intent, target companies, or other contextual data to third-party services and may cause compliance issues in restricted environments.

Ssd 3

Medium

Confidence: 97% confidence
Finding: The cross-session memory design instructs persistent collection of sensitive candidate data, interview records, weaknesses, and story-bank content in a shared memory file. In context, this is especially risky because job-search records can contain personally identifiable information, employment history, and sensitive performance notes that should not be retained indefinitely or broadly accessible.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.