Back to skill

Security audit

全网招标采购与供应商寻源-必联网

Security checks for vulnerabilities and agentic risk

Overview

This procurement assistant needs review because optional signup sends a reusable device identifier while describing it as non-identifying.

Prefer a manually obtained API key to skip device fingerprinting. Before accepting automatic signup, ask how the provider retains and deletes the reusable device identifier. Business search terms go to Zhiliaobiaoxun, queries can consume credits, and displayed update notices trigger an authenticated acknowledgement. Protect the saved key and generated login links. This review inspected the bundled instructions; server-side privacy practices were not verified.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (10)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The declared purpose concerns procurement and supplier intelligence, whereas this chunk accesses local device information and derives a device fingerprint. That capability is not disclosed or evidently necessary for the stated purpose. Although the file is named auto-register.md, the supplied code does not itself demonstrate registration, transmission, or storage of the resulting hash.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied code performs device identification rather than procurement or supplier research. Hardware-address collection and hashing are not disclosed in the description, and the chunk shows no connection to a necessary supporting procurement operation. It does not show transmission, storage, or actual registration of the resulting identifier.

Content

No source excerpt is available for this finding.

The onboarding flow collects device fingerprints, a capability not justified by tender search or supplier analysis.

Medium
Category
Excessive Agency
Confidence
88% confidence
Finding

L052–L056 instruct the agent to collect platform, architecture, and a MAC-address hash for remote account registration when credentials are absent. Although user consent is required, device fingerprinting is not an obvious requirement of the manifest's procurement-data queries and supplier-analysis functions, and the manifest does not declare this capability.

Content

No source excerpt is available for this finding.

The query workflow sends user-supplied business search details to a third-party service without a user-facing privacy disclosure.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The company-analysis workflow instructs the agent to submit company identifiers through several API calls to the external service identified at L019. Although the file documents the API mechanics, it does not instruct the agent to warn users that their company names, procurement keywords, and other query parameters will leave the conversation, potentially revealing confidential research or sourcing interests. The consent requirement at L052–L055 covers device-feature collection for registration, not subsequent business queries.

Content

No source excerpt is available for this finding.

Automatic company searches transmit user-provided company names to an external API without a user-facing disclosure.

Medium
Category
Output Handling
Confidence
87% confidence
Finding

The workflow instructs the agent to send the company name from the user's request to search_company and use the matched companies in subsequent queries without interrupting the user. Although the API operations are documented, there is no user-facing warning that these searches disclose the user's company research targets to the external service; this is an instructed workflow, not merely a tool listing.

Content

No source excerpt is available for this finding.

The skill sends an authenticated notice-viewing receipt without informing the user about this reporting.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The workflow instructs the agent to POST a notice identifier and displayed: true to the external service after showing an update notice. This reports the user's notice-display status and changes server-side acknowledgement state, but the file provides no user-facing disclosure and explicitly says not to show the technical receipt.

Content

No source excerpt is available for this finding.

The skill collects and transmits a persistent hardware-derived identifier for trial-account deduplication, beyond its stated procurement-search purpose.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

This source transmits a deterministic, unsalted MAC-derived identifier while its privacy explanation and consent template describe the collected features as having no identity meaning and collecting no personal information. Hashing does not eliminate persistent device linkability or prevent testing candidate MAC addresses, so those assurances undermine informed consent; the explicit opt-in and manual alternative reduce, but do not remove, this privacy risk.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% 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 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

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

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**