Back to skill

Security audit

精准寻标与获客引擎-寻标宝

Security checks for vulnerabilities and agentic risk

Overview

The tender-search features fit their purpose, but misleading device-privacy assurances and conflicting account-registration rules warrant review before installation.

Review the device-registration flow before installing. Prefer a manually obtained API key if you want to avoid hardware-derived identification, and require explicit confirmation before changing an existing account or key. Ask the publisher to correct the privacy assurances and reconcile credential-recovery instructions. Queries consume credits, and login links grant temporary account access. Provider-side fingerprint retention and deletion could not be established from the artifacts.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied code derives a local device identifier rather than querying tender data or identifying business prospects. Hardware-identifier collection is not declared or evidently necessary to the stated customer-acquisition purpose. The code does not show transmission of the resulting hash.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The code accesses local system networking metadata and generates a device fingerprint rather than querying bidding data or identifying business prospects. Device identification or registration is not disclosed in the declared purpose, and the supplied chunk shows no connection to the customer-discovery workflow. No transmission or exfiltration is shown.

Content

No source excerpt is available for this finding.

Authentication-error handling directs automatic registration even when configured credentials exist, contradicting the registration boundary.

Medium
Category
Excessive Agency
Confidence
95% confidence
Finding

The authentication-error table directs registration even when an existing key is configured, contradicting the explicit registration gate. An agent following this recovery path could unnecessarily collect device features, create a different account, or replace locally stored credentials instead of repairing authentication; the earlier consent requirement mitigates but does not resolve the contradictory instructions.

Content

No source excerpt is available for this finding.

The skill mandates an unrelated package-update notification and acknowledgment workflow.

Medium
Category
Excessive Agency
Confidence
93% confidence
Finding

L334–L336 instruct the agent to check for package upgrades through an authenticated account endpoint, relay a server-provided update message verbatim, and submit a POST acknowledgment after displaying it. This package-maintenance and notification-tracking workflow is not a direct requirement of tender-based customer discovery or customer/supplier analysis, and the manifest does not declare it.

Content

No source excerpt is available for this finding.

The skill collects and transmits a persistent hardware fingerprint, beyond its stated tender-sourcing purpose.

Medium
Category
Excessive Agency
Confidence
90% confidence
Finding

The registration flow transmits a deterministic, unsalted MAC-derived identifier while describing the collected features as having no identity meaning and collecting no personal information. Hashing a MAC does not make it anonymous: it remains linkable and can permit candidate matching, so the consent language materially understates the privacy implications even though collection is opt-in.

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**

L351–L353 mandate a package-update notification and acknowledgement workflow unrelated to tender sourcing.

Low
Category
Excessive Agency
Confidence
86% confidence
Finding

The manifest describes identifying sales prospects and analyzing customer/supplier relationships through tender data. These instructions additionally require a session-level package-update check and an authenticated POST recording that an update notice was displayed, introducing remote maintenance-state changes that are not needed for the declared sourcing workflow.

Content

No source excerpt is available for this finding.