Back to skill

Security audit

push-server

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed notification and webhook skill with meaningful data-sharing and channel-management powers, but the artifacts align those powers with the stated purpose and include user-control guidance.

Install only if you want the agent to send task results to push-server and your configured notification or webhook endpoints. Treat the channel code like a credential, do not push secrets or private tokens in message bodies, and confirm carefully before channel delete, disable, or reset operations.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (18)

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly supports pushing scraped or structured content to external webhook or business servers, but the top-level description does not prominently warn users that data may be transmitted off-platform to third-party endpoints. This can lead to accidental exfiltration of sensitive task results, especially because notification delivery is the core function of the skill and users may not realize external delivery is occurring.

Content

No source excerpt is available for this finding.

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 323)May include surrounding context.

}

text

| 字段                                                             | 说明                                                                                                                                            |
|----------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------|
| `code` / `message`                                             | 0 表示成功,其余为业务错误码 / 错误描述                                                                                                                        |
| `data.channelCode` / `channelName` / `pushType` / `createTime` | 父通道(组)信息                                                                                                                                      |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 325)May include surrounding context.

md
| 字段                                                             | 说明                                                                                                                                            |
|----------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------|
| `code` / `message`                                             | 0 表示成功,其余为业务错误码 / 错误描述                                                                                                                        |
| `data.channelCode` / `channelName` / `pushType` / `createTime` | 父通道(组)信息                                                                                                                                      |
| `data.rows[].channelName`                                      | 通道名称                                                                                                                                          |
| `data.rows[].pushType`                                         | 推送类型编码,见 5.7 枚举表                                                                                                                              |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 326)May include surrounding context.

md
| 字段                                                             | 说明                                                                                                                                            |
|----------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------|
| `code` / `message`                                             | 0 表示成功,其余为业务错误码 / 错误描述                                                                                                                        |
| `data.channelCode` / `channelName` / `pushType` / `createTime` | 父通道(组)信息                                                                                                                                      |
| `data.rows[].channelName`                                      | 通道名称                                                                                                                                          |
| `data.rows[].pushType`                                         | 推送类型编码,见 5.7 枚举表                                                                                                                              |
| `data.rows[].pushTypeName`                                     | 推送类型标识(枚举名小写,如 `website`、`web_hook_push`)                                                                                                     |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 327)May include surrounding context.

md
|----------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------|
| `code` / `message`                                             | 0 表示成功,其余为业务错误码 / 错误描述                                                                                                                        |
| `data.channelCode` / `channelName` / `pushType` / `createTime` | 父通道(组)信息                                                                                                                                      |
| `data.rows[].channelName`                                      | 通道名称                                                                                                                                          |
| `data.rows[].pushType`                                         | 推送类型编码,见 5.7 枚举表                                                                                                                              |
| `data.rows[].pushTypeName`                                     | 推送类型标识(枚举名小写,如 `website`、`web_hook_push`)                                                                                                     |
| `data.rows[].pushTypeDesc`                                     | 推送类型描述(如「浏览器」「webhook推送」)                                                                                                                     |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 328)May include surrounding context.

md
| `code` / `message`                                             | 0 表示成功,其余为业务错误码 / 错误描述                                                                                                                        |
| `data.channelCode` / `channelName` / `pushType` / `createTime` | 父通道(组)信息                                                                                                                                      |
| `data.rows[].channelName`                                      | 通道名称                                                                                                                                          |
| `data.rows[].pushType`                                         | 推送类型编码,见 5.7 枚举表                                                                                                                              |
| `data.rows[].pushTypeName`                                     | 推送类型标识(枚举名小写,如 `website`、`web_hook_push`)                                                                                                     |
| `data.rows[].pushTypeDesc`                                     | 推送类型描述(如「浏览器」「webhook推送」)                                                                                                                     |
| `data.rows[].channelCode`                                      | 该子通道的 32 位通道码:可直接当 MCP 凭证用、**是「当前通道怎么找」的比对依据**、也是 5.3/5.4/5.5 定位目标通道的唯一入参                                                                     |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 329)May include surrounding context.

md
| `data.channelCode` / `channelName` / `pushType` / `createTime` | 父通道(组)信息                                                                                                                                      |
| `data.rows[].channelName`                                      | 通道名称                                                                                                                                          |
| `data.rows[].pushType`                                         | 推送类型编码,见 5.7 枚举表                                                                                                                              |
| `data.rows[].pushTypeName`                                     | 推送类型标识(枚举名小写,如 `website`、`web_hook_push`)                                                                                                     |
| `data.rows[].pushTypeDesc`                                     | 推送类型描述(如「浏览器」「webhook推送」)                                                                                                                     |
| `data.rows[].channelCode`                                      | 该子通道的 32 位通道码:可直接当 MCP 凭证用、**是「当前通道怎么找」的比对依据**、也是 5.3/5.4/5.5 定位目标通道的唯一入参                                                                     |
| `data.rows[].webhookUrl`                                       | webhook 通道的接收地址,非 webhook 类型为空                                                                                                                |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 330)May include surrounding context.

md
| `data.rows[].channelName`                                      | 通道名称                                                                                                                                          |
| `data.rows[].pushType`                                         | 推送类型编码,见 5.7 枚举表                                                                                                                              |
| `data.rows[].pushTypeName`                                     | 推送类型标识(枚举名小写,如 `website`、`web_hook_push`)                                                                                                     |
| `data.rows[].pushTypeDesc`                                     | 推送类型描述(如「浏览器」「webhook推送」)                                                                                                                     |
| `data.rows[].channelCode`                                      | 该子通道的 32 位通道码:可直接当 MCP 凭证用、**是「当前通道怎么找」的比对依据**、也是 5.3/5.4/5.5 定位目标通道的唯一入参                                                                     |
| `data.rows[].webhookUrl`                                       | webhook 通道的接收地址,非 webhook 类型为空                                                                                                                |
| `data.rows[].signSecret`                                      | 钉钉 / 企业微信 / 飞书群机器人的加签 Secret,未启用加签或非这三类群机器人时为空                                                                                                  |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 332)May include surrounding context.

md
| `data.rows[].pushTypeName`                                     | 推送类型标识(枚举名小写,如 `website`、`web_hook_push`)                                                                                                     |
| `data.rows[].pushTypeDesc`                                     | 推送类型描述(如「浏览器」「webhook推送」)                                                                                                                     |
| `data.rows[].channelCode`                                      | 该子通道的 32 位通道码:可直接当 MCP 凭证用、**是「当前通道怎么找」的比对依据**、也是 5.3/5.4/5.5 定位目标通道的唯一入参                                                                     |
| `data.rows[].webhookUrl`                                       | webhook 通道的接收地址,非 webhook 类型为空                                                                                                                |
| `data.rows[].signSecret`                                      | 钉钉 / 企业微信 / 飞书群机器人的加签 Secret,未启用加签或非这三类群机器人时为空                                                                                                  |
| `data.rows[].status`                                           | 通道状态(`1` 已启用 / `0` 未启用,见 5.8)                                                                                                                 |
| `data.rows[].channelMemberRelId`                               | **该子通道创建人的成员 ID**,`nickname` 同属这条成员记录;**成员接口入参的唯一来源**,可直接作为 5.6 的入参。**仅 `pushType=1`(浏览器)的子通道有值**,其余为 `null`(顶层 `data` 不带成员字段,也没有成员分页接口可查其它人) |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 333)May include surrounding context.

md
| `data.rows[].pushTypeDesc`                                     | 推送类型描述(如「浏览器」「webhook推送」)                                                                                                                     |
| `data.rows[].channelCode`                                      | 该子通道的 32 位通道码:可直接当 MCP 凭证用、**是「当前通道怎么找」的比对依据**、也是 5.3/5.4/5.5 定位目标通道的唯一入参                                                                     |
| `data.rows[].webhookUrl`                                       | webhook 通道的接收地址,非 webhook 类型为空                                                                                                                |
| `data.rows[].signSecret`                                      | 钉钉 / 企业微信 / 飞书群机器人的加签 Secret,未启用加签或非这三类群机器人时为空                                                                                                  |
| `data.rows[].status`                                           | 通道状态(`1` 已启用 / `0` 未启用,见 5.8)                                                                                                                 |
| `data.rows[].channelMemberRelId`                               | **该子通道创建人的成员 ID**,`nickname` 同属这条成员记录;**成员接口入参的唯一来源**,可直接作为 5.6 的入参。**仅 `pushType=1`(浏览器)的子通道有值**,其余为 `null`(顶层 `data` 不带成员字段,也没有成员分页接口可查其它人) |
| `data.rows[].pushUrl`                                          | website 通道的分享/管理页面地址(share.html),打开即可查看与管理该通道,非 website 类型为空                                                                                                   |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 334)May include surrounding context.

md
| `data.rows[].channelCode`                                      | 该子通道的 32 位通道码:可直接当 MCP 凭证用、**是「当前通道怎么找」的比对依据**、也是 5.3/5.4/5.5 定位目标通道的唯一入参                                                                     |
| `data.rows[].webhookUrl`                                       | webhook 通道的接收地址,非 webhook 类型为空                                                                                                                |
| `data.rows[].signSecret`                                      | 钉钉 / 企业微信 / 飞书群机器人的加签 Secret,未启用加签或非这三类群机器人时为空                                                                                                  |
| `data.rows[].status`                                           | 通道状态(`1` 已启用 / `0` 未启用,见 5.8)                                                                                                                 |
| `data.rows[].channelMemberRelId`                               | **该子通道创建人的成员 ID**,`nickname` 同属这条成员记录;**成员接口入参的唯一来源**,可直接作为 5.6 的入参。**仅 `pushType=1`(浏览器)的子通道有值**,其余为 `null`(顶层 `data` 不带成员字段,也没有成员分页接口可查其它人) |
| `data.rows[].pushUrl`                                          | website 通道的分享/管理页面地址(share.html),打开即可查看与管理该通道,非 website 类型为空                                                                                                   |
| `data.rows[].qrCodeUrl`                                        | **推送记录页面**的二维码,扫码即可在手机上查看推送记录,非 website 类型为空                                                                                                 |

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 336)May include surrounding context.

md
| `data.rows[].signSecret`                                      | 钉钉 / 企业微信 / 飞书群机器人的加签 Secret,未启用加签或非这三类群机器人时为空                                                                                                  |
| `data.rows[].status`                                           | 通道状态(`1` 已启用 / `0` 未启用,见 5.8)                                                                                                                 |
| `data.rows[].channelMemberRelId`                               | **该子通道创建人的成员 ID**,`nickname` 同属这条成员记录;**成员接口入参的唯一来源**,可直接作为 5.6 的入参。**仅 `pushType=1`(浏览器)的子通道有值**,其余为 `null`(顶层 `data` 不带成员字段,也没有成员分页接口可查其它人) |
| `data.rows[].pushUrl`                                          | website 通道的分享/管理页面地址(share.html),打开即可查看与管理该通道,非 website 类型为空                                                                                                   |
| `data.rows[].qrCodeUrl`                                        | **推送记录页面**的二维码,扫码即可在手机上查看推送记录,非 website 类型为空                                                                                                 |

#### 通道码的层级与含义

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · references/api.md (reported line 337)May include surrounding context.

md
| `data.rows[].status`                                           | 通道状态(`1` 已启用 / `0` 未启用,见 5.8)                                                                                                                 |
| `data.rows[].channelMemberRelId`                               | **该子通道创建人的成员 ID**,`nickname` 同属这条成员记录;**成员接口入参的唯一来源**,可直接作为 5.6 的入参。**仅 `pushType=1`(浏览器)的子通道有值**,其余为 `null`(顶层 `data` 不带成员字段,也没有成员分页接口可查其它人) |
| `data.rows[].pushUrl`                                          | website 通道的分享/管理页面地址(share.html),打开即可查看与管理该通道,非 website 类型为空                                                                                                   |
| `data.rows[].qrCodeUrl`                                        | **推送记录页面**的二维码,扫码即可在手机上查看推送记录,非 website 类型为空                                                                                                 |

#### 通道码的层级与含义

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

This markdown reference is entirely written in Chinese and presents mandatory usage guidance without indicating that users may choose another language or locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly documented and justified, which it is not here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The display name and the main body of the skill document are presented in Chinese, which can impose a language choice on users reading the primary instructions. Although English metadata fields exist, the operational guidance in this file does not explicitly offer users a choice of language or indicate that Chinese is optional.

Content

No source excerpt is available for this finding.