Security audit

Company search fengniao

Security checks across malware telemetry and agentic risk

Overview

This is a coherent Riskbird company-lookup skill, with disclosed external API calls and credential handling, but users should understand that lookup terms and API keys are sent in URL parameters.

Install only if you are comfortable sending company names, person-name lookup terms, entid values, and any configured FN_API_KEY to Riskbird. Prefer a dedicated API key, avoid printing the full key in terminals or logs, and be aware that broad business-background prompts may invoke this skill.

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 (9)

Missing User Warnings

Low

Confidence: 84% confidence
Finding: The setup guide tells users to display the full `FN_API_KEY` in the terminal to verify configuration. While this is common troubleshooting guidance, it unnecessarily exposes a live secret to anyone observing the screen or reviewing terminal recordings/log capture, increasing the chance of credential disclosure.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The skill is configured with auto_invoke and very broad example triggers such as generic company lookup, due diligence, and risk screening requests. This can cause the skill to activate on loosely related prompts and send company names or person names to an external service without sufficiently explicit user intent or confirmation, especially in ambiguous cases.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The skill states that requests are sent using an API key in the URL parameter apikey and does not present a prominent user warning that lookup terms are transmitted to an external provider. URL query parameters are commonly logged by clients, proxies, servers, and observability tooling, which increases exposure of both credentials and searched entities.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The code sends the API key as a URL query parameter and forwards user-supplied parameters directly to a remote service. Query-string credentials are commonly exposed in logs, browser history, proxies, and monitoring systems, which increases the chance of credential leakage; additionally, there is no validation or user-consent/disclosure layer before transmitting potentially sensitive company-search inputs off-device. In this skill context, users are likely to submit business due-diligence targets and risk-related identifiers, so silent transmission to a third-party API is materially privacy-relevant.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The file hardcodes a shared built-in API key and silently falls back to it whenever a user-supplied key is absent. This exposes a reusable credential to anyone with access to the skill source, enabling unauthorized third-party use, quota exhaustion, and possible attribution of abusive traffic to the skill author or shared account.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: Aliases such as “背景调查”, “对方背景”, “签合同”, “供应商”, “面试前”, and “目标公司” are broad natural-language phrases that can appear in ordinary conversation without an explicit intent to invoke this skill. That increases the chance of accidental routing to a company-risk lookup flow, potentially causing unintended external queries or disclosure of queried entities in contexts where the user was only speaking generally.

Vague Triggers

Medium

Confidence: 92% confidence
Finding: Single-word aliases like “法人”, “官网”, “电话”, and “邮箱” are highly underspecified and likely to collide with many unrelated user requests. In an assistant environment, these generic tokens can trigger the skill unexpectedly, leading to unintended company-information lookups and possible privacy, logging, or workflow-integrity issues.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: Keywords like “地址” and “法人” are highly generic and overlap with ordinary conversation, making accidental invocation more likely than with domain-specific terminology. In this skill, accidental invocation is more concerning because the tool can return business registry, executive, and risk information tied to real entities, so weak trigger precision expands privacy and data-minimization risk.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: Keywords like “地址” and “法人” are highly generic and overlap with ordinary conversation, making accidental invocation more likely than with domain-specific terminology. In this skill, accidental invocation is more concerning because the tool can return business registry, executive, and risk information tied to real entities, so weak trigger precision expands privacy and data-minimization risk.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.