Back to skill

Security audit

拟建项目跟踪-立项审批阶段就发现

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its business-opportunity search purpose, but it stores credentials locally, can register an account using device-derived data, and emits signed access links in shareable reports.

Review before installing. Use a pre-provisioned API key if possible, avoid auto-registration if you do not want device-derived data sent to the service, and do not publicly share generated reports unless signed sk links have been removed or you are comfortable sharing that access. Confirm where ~/.zlbx/config.json will be stored and how to delete it later.

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)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

声明描述的是一个面向用户的商机发现/查询型技能,重点在数据检索、筛选、跟踪和机会挖掘;而提供的代码块只是下游展示层组件,负责把现成 JSON 数据渲染为品牌化 HTML 页面。虽然报告内容字段涵盖‘拟建/采购意向/临期续约’等板块,与声明领域相关,但代码本身没有任何获取数据、调用外部资源、搜索项目、计算排序或筛选状态的行为。因此其实际主功能与声明的核心能力存在明显落差。代码没有发现额外敏感或越权行为;问题在于声明的是“商机发现助手”,代码实际是“报告生成器/导出器”。

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The activation guidance is extremely broad, instructing the agent to invoke this skill for many loosely related 'project research' or early-opportunity requests, even when the user does not mention the core concept directly. Over-broad routing can cause unnecessary transmission of user queries to an external service, surprise the user with unintended API usage or billing, and expand the situations in which local report generation and remote account workflows are triggered.

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 81)May include surrounding context.

md
完整模式下,对话清单输出后默认再用 `scripts/render_report.py` 生成一份可分享的 HTML 版商机清单并告知保存路径(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 85)May include surrounding context.

md
完整模式下,对话清单输出后默认再用 `scripts/render_report.py` 生成一份可分享的 HTML 版商机清单并告知保存路径(详见 report-template.md「HTML 报告导出」)。

Hidden Instructions

High
Category
Prompt Injection
Confidence
96% confidence
Finding

This duplicate finding points to the same hidden instruction block, which materially changes request behavior by requiring package identity headers and a follow-up acknowledgement action after displaying update notices. Because the logic is concealed in comments, it increases the risk of undisclosed metadata transmission and complicates security review, especially in a skill that already handles API keys, account state, and local file output.

Content

Scanner excerpt · SKILL.md (reported line 171)May include surrounding context.

  • 用户想持续跟踪 → 介绍订阅模式(见上方「订阅模式」)
  • 以上都不贴切 → 询问是否要对某条高价值商机深挖,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: proposed-project-tracker/1.0.8

Hidden Instructions

High
Category
Prompt Injection
Confidence
96% confidence
Finding

This duplicate finding points to the same hidden instruction block, which materially changes request behavior by requiring package identity headers and a follow-up acknowledgement action after displaying update notices. Because the logic is concealed in comments, it increases the risk of undisclosed metadata transmission and complicates security review, especially in a skill that already handles API keys, account state, and local file output.

Content

Scanner excerpt · SKILL.md (reported line 171)May include surrounding context.

  • 用户想持续跟踪 → 介绍订阅模式(见上方「订阅模式」)
  • 以上都不贴切 → 询问是否要对某条高价值商机深挖,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: proposed-project-tracker/1.0.8

Hidden Instructions

High
Category
Prompt Injection
Confidence
89% confidence
Finding

This duplicate hidden-instruction finding points to the same concealed identity-header directive at the top of the file. Concealed network-influencing instructions are inappropriate in a content skill because they can bypass user understanding and complicate auditing of outbound data flows.

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: opportunity-radar/1.0.3

Hidden Instructions

High
Category
Prompt Injection
Confidence
89% confidence
Finding

This duplicate hidden-instruction finding points to the same concealed identity-header directive at the top of the file. Concealed network-influencing instructions are inappropriate in a content skill because they can bypass user understanding and complicate auditing of outbound data flows.

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: opportunity-radar/1.0.3

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The file embeds a full account lifecycle for an unrelated project-tracking skill: API key discovery, automatic account registration, key persistence, balance checks, and recharge/login flows. This expands the skill’s authority far beyond its declared business purpose and creates unnecessary access to local state, credentials, and third-party account operations, increasing the attack surface for data exfiltration and unauthorized account provisioning.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The instructions direct the agent to inspect environment variables and local config files for API keys and, absent those, collect device fingerprints including OS, architecture, and a MAC-derived hash. For a proposed-project tracking skill, this collection is not functionally necessary and enables host reconnaissance plus stable device correlation, which is sensitive even if the raw MAC is hashed.

Content

No source excerpt is available for this finding.

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: opportunity-radar/1.0.3

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/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: opportunity-radar/1.0.3

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: opportunity-radar/1.0.3

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/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: opportunity-radar/1.0.3

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The template requires preserving full API-returned URLs containing an sk login-bypass parameter and presenting them directly to users. If these links are shared, logged, or embedded in exports, the sk token may grant unintended access to protected resources, effectively turning the report into a bearer-token distribution channel.

Content

No source excerpt is available for this finding.

Ssd 3

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill mandates exposing full authenticated links with sk parameters both in chat and in exported HTML/JSON artifacts. This amplifies the risk of credential leakage because the token can propagate through user sharing, browser history, logs, screenshots, exported files, or downstream systems that treat the link as ordinary content.

Content

No source excerpt is available for this finding.

Hidden Instructions

High
Category
Prompt Injection
Confidence
96% confidence
Finding

This duplicate hidden-instruction finding reflects the same underlying issue: concealed directives at the top of the file instruct the agent to send fixed identity headers and use configured credentials. In the context of a skill that makes external API calls, hidden instructions materially increase the risk of undisclosed data transmission and policy bypass.

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: opportunity-radar/1.0.3

Hidden Instructions

High
Category
Prompt Injection
Confidence
96% confidence
Finding

This duplicate hidden-instruction finding reflects the same underlying issue: concealed directives at the top of the file instruct the agent to send fixed identity headers and use configured credentials. In the context of a skill that makes external API calls, hidden instructions materially increase the risk of undisclosed data transmission and policy bypass.

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: opportunity-radar/1.0.3

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 64)May include surrounding context.

python
ROUTE_COLOR = {"拟建": "#7c5cbf", "意向": "#0b7a6a", "临期": "#b9770e"}

# 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
90% confidence
Finding

The skill explicitly performs local file output by writing HTML reports to ~/zlbx-opportunity-radar-files/, but it declares no permissions or allowed-tools scope to bound that capability. That creates an authorization gap: an integrator or user cannot easily tell that the skill can write to disk, and a modified workflow could abuse the same undeclared capability to persist unexpected data or artifacts.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The file is entirely written to operate in Chinese and instructs the assistant with mandatory behavior in that locale, but does not indicate that users may choose another language. Under the policy, forcing a specific language without user opt-in is a natural-language locale violation unless the constraint is explicitly justified as region-specific.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
91% confidence
Finding

This section explicitly instructs transmission of locally collected device features to an external registration endpoint. External transmission of host-derived identifiers is security-relevant because it exports local metadata off-device, and here it occurs in a skill whose stated purpose does not justify such collection or sharing.

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
96% confidence
Finding

This duplicate persistence finding refers to the same instruction set for creating ~/.zlbx, merging config, and storing the API key plus source metadata. Predictable-file credential persistence is especially dangerous here because it is paired with automatic registration and continued authenticated use, creating a stealthy long-lived authorization footprint.

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**

Session Persistence

Medium
Category
Rogue Agent
Confidence
96% confidence
Finding

This duplicate persistence finding refers to the same instruction set for creating ~/.zlbx, merging config, and storing the API key plus source metadata. Predictable-file credential persistence is especially dangerous here because it is paired with automatic registration and continued authenticated use, creating a stealthy long-lived authorization footprint.

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**