Back to skill

Security audit

企业背调助手-给个公司名出份报告

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its company-background-check purpose, but it needs review because it manages credentials, collects a device-derived identifier for registration, and preserves shareable signed links.

Install only if you are comfortable with this skill using a Zhiliaobiaoxun API key or creating one after consent, storing that key in ~/.zlbx/config.json, consuming paid or trial credits, and generating reports that may contain signed sk links and returned contact information. Avoid sharing generated HTML reports outside the intended audience unless you are comfortable exposing those links.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
Findings (34)

Vague Triggers

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The activation logic is intentionally expansive: it says the skill must be used whenever a user wants to understand a company, even if they did not ask for background checks. This can hijack unrelated requests, trigger unnecessary external queries, registration flows, or spending, and override more appropriate or safer skills.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 89)May include surrounding context.

md
默认再用 `scripts/render_report.py` 生成一份可分享的 HTML 版报告并告知保存路径(详见 report-template.md「HTML 报告导出」)。

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 93)May include surrounding context.

md
默认再用 `scripts/render_report.py` 生成一份可分享的 HTML 版报告并告知保存路径(详见 report-template.md「HTML 报告导出」)。

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 · SKILL.md (reported line 185)May include surrounding context.

  • 用户想持续跟踪该公司 → 介绍竞对动态周报(见上方「竞对动态周报」)
  • 以上都不贴切 → 询问是否要对某个竞对做双公司对比,或引导访问知了商机大师 https://agent.zhiliaobiaoxun.com

安装包身份与升级检查:所有发往知了标讯 API 的请求(包括账户查询)必须额外携带以下两个请求头:

text
X-Skill-Package: company-background-check/1.0.6

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 · SKILL.md (reported line 185)May include surrounding context.

  • 用户想持续跟踪该公司 → 介绍竞对动态周报(见上方「竞对动态周报」)
  • 以上都不贴切 → 询问是否要对某个竞对做双公司对比,或引导访问知了商机大师 https://agent.zhiliaobiaoxun.com

安装包身份与升级检查:所有发往知了标讯 API 的请求(包括账户查询)必须额外携带以下两个请求头:

text
X-Skill-Package: company-background-check/1.0.6

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 · references/api-quick.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

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 · references/api-quick.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Hidden Instructions

High
Category
Prompt Injection
Confidence
95% confidence
Finding

The skill embeds hidden operational instructions in comments that direct request identity headers and package/source attribution. Hidden control instructions are dangerous because they can steer agent behavior outside the visible user intent and can be used to manipulate outbound requests or bypass normal trust boundaries.

Content

Scanner excerpt · references/auto-register.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Hidden Instructions

High
Category
Prompt Injection
Confidence
95% confidence
Finding

The skill embeds hidden operational instructions in comments that direct request identity headers and package/source attribution. Hidden control instructions are dangerous because they can steer agent behavior outside the visible user intent and can be used to manipulate outbound requests or bypass normal trust boundaries.

Content

Scanner excerpt · references/auto-register.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Hidden Instructions

High
Category
Prompt Injection
Confidence
91% confidence
Finding

This duplicate finding points to the same hidden instruction block embedding request-identity directives inside comments. Such concealed instructions are dangerous because they mix untrusted content with execution-affecting behavior, making it easier for downstream agents to obey sensitive directives that were not meant as ordinary display text.

Content

Scanner excerpt · references/report-template.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Hidden Instructions

High
Category
Prompt Injection
Confidence
91% confidence
Finding

This duplicate finding points to the same hidden instruction block embedding request-identity directives inside comments. Such concealed instructions are dangerous because they mix untrusted content with execution-affecting behavior, making it easier for downstream agents to obey sensitive directives that were not meant as ordinary display text.

Content

Scanner excerpt · references/report-template.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The file repeatedly mandates preserving and redistributing original URLs that contain sk login-bypass parameters, and even emphasizes not removing them. Propagating bearer-style access tokens in user-visible output can leak privileged access, enable unauthorized sharing, and expose private or paid resources beyond the intended recipient.

Content

No source excerpt is available for this finding.

Hidden Instructions

High
Category
Prompt Injection
Confidence
96% confidence
Finding

This is the same hidden-instructions issue in duplicate form: the workflow includes concealed directives telling the agent which identity headers to send with authenticated requests. Such hidden control text is risky because it can silently alter how credentials are presented to upstream services and is especially concerning in a skill that already instructs use of an existing API key.

Content

Scanner excerpt · references/workflow.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Hidden Instructions

High
Category
Prompt Injection
Confidence
96% confidence
Finding

This is the same hidden-instructions issue in duplicate form: the workflow includes concealed directives telling the agent which identity headers to send with authenticated requests. Such hidden control text is risky because it can silently alter how credentials are presented to upstream services and is especially concerning in a skill that already instructs use of an existing API key.

Content

Scanner excerpt · references/workflow.md (reported line 1)May include surrounding context.

请求身份:使用已配置的 API Key,并保留本包的 X-Client,携带下列完整身份头;不得沿用其它包的版本或来源。

text
X-Client: company-intel/1.0.2

Obfuscated Code

High
Category
Supply Chain
Confidence
50% confidence
Finding

Code contains obfuscation (base64, hex encoding with execution). This is often used to hide malicious functionality.

Content

Scanner excerpt · scripts/render_report.py (reported line 93)May include surrounding context.

python
# 知了标讯白色 logo(299x96 PNG base64 内嵌,保证报告离线/打印/转发时不裂图)
_LOGO_B64 = "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
...[truncated 27 chars]

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding

The skill explicitly writes HTML reports to a local directory but does not declare any tool scope, permissions, or allowed-tools boundary for file writing. Undeclared write capability increases the chance of overbroad filesystem access or misuse by the hosting agent, especially when processing untrusted company names that may influence output filenames or paths.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The skill description and all prescribed user-facing phrasing are written as mandatory Chinese output patterns, but there is no indication that users may choose another language or that the locale restriction is required. This can violate language/locale policy when used in broader multilingual environments.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
82% confidence
Finding

The file instructs the skill to send an API key and fixed identity headers on every request to an external API, which is a network operation involving credentials. It describes the mechanics but provides no user-facing warning that requests will transmit configured credentials and service-identifying metadata to a third-party endpoint.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill explicitly supports retrieving and displaying project contact phone numbers, including full numbers for paid accounts, but provides no user-facing privacy, consent, retention, or permissible-use warning. Because this skill is designed for company background checks, normalizing access to personal contact data increases the risk of privacy misuse, unauthorized outreach, and data handling noncompliance.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The document adds account provisioning, device fingerprint collection, credential storage, and recharge/login handling to a skill whose stated purpose is company background investigation. This is an unjustified expansion of scope that increases attack surface and enables sensitive local and network operations unrelated to the user’s expected task.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill instructs collection of device fingerprints (platform, architecture, MAC-derived hash) and ties them to account creation, even though those identifiers are not necessary for generating a company background report. Collecting and transmitting host-derived identifiers creates privacy risk and device tracking capability beyond the skill’s declared purpose.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

This section documents sending collected device features to an external auto-registration endpoint. In context, the transmission includes host-derived identifiers for account creation, making it a real data exfiltration/privacy concern rather than a generic network call.

Content

Scanner excerpt · references/auto-register.md (reported line 130)May include surrounding context.

md
> ### ⚠️ 请求体必须用 JSON 序列化函数生成,不要手拼字符串
>
> 用 `json.dumps(payload)` / `requests.post(url, json=payload)` / `JSON.stringify(payload)`,
> 或 `curl -d @file`;**不要用字符串拼接,也不要用 Python 的 `str(dict)`**
> (后者产出单引号,服务端会报 `Expecting property name enclosed in double quotes`)。
>

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/auto-register.md (reported line 131)May include surrounding context.

md
> ### ⚠️ 请求体必须用 JSON 序列化函数生成,不要手拼字符串
>
> 用 `json.dumps(payload)` / `requests.post(url, json=payload)` / `JSON.stringify(payload)`,
> 或 `curl -d @file`;**不要用字符串拼接,也不要用 Python 的 `str(dict)`**
> (后者产出单引号,服务端会报 `Expecting property name enclosed in double quotes`)。
>
> 历史教训:曾有版本采集 `home_path`,Windows 的 `C:\Users\alice` 直接拼进 JSON 字符串时

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill directs the agent to create and modify ~/.zlbx/config.json and persist API keys for future sessions, which is unrelated to a read/report-oriented due-diligence function. Persisting credentials from within a content skill broadens privilege, creates long-lived secrets on disk, and can surprise users who only expected a one-time report.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
97% confidence
Finding

The skill explicitly creates a persistent config directory/file and stores API keys for reuse across sessions, then instructs immediate use of the stored key. This introduces session persistence and durable credential storage from a skill that should not manage local authentication state, increasing exposure if the host or account is later compromised.

Content

Scanner excerpt · references/auto-register.md (reported line 195)May include surrounding context.

text

注意事项:
- 目录不存在时先 `mkdir -p ~/.zlbx`
- 文件已存在时**合并而非覆盖**(保留用户可能的其他配置)
- `source: "auto"` 字段必须写入,**这是后续判断「是否输出自动登录链接」的关键依据**