Back to skill

Security audit

Ai Docgen Expert

Security checks for vulnerabilities and agentic risk

Overview

The skill is marked suspicious because its workflow reportedly runs documentation examples without clear isolation, while several referenced files were only partly inspected.

Consider postponing installation until the example-running workflow has clear safeguards, such as an isolated environment without credentials or network access, and the partially inspected references can be reviewed. The verdict is suspicious (Review), with medium overall confidence; the scan does not establish malicious intent, and its evidence is incomplete.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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
Findings (12)

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 15)May include surrounding context.

md
| kb/doc-taxonomy.md | Diátaxis 四象限判型、混写崩塌(blur)拆分、README 定位、读者-时刻矩阵、信息架构(DT-1~12) | 动笔前判型 / 文档体系设计 |

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 32)May include surrounding context.

md
- 判型(tutorial/how-to/reference/explanation)/ 混写拆分 / README / 信息架构 → `kb/doc-taxonomy.md`

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 17)May include surrounding context.

md
| kb/code-comment.md | 注释价值判据(why/context)、Google 官方谱系、噪音注释反模式、TODO 纪律、语言风格差异(CC-1~12) | 写注释/docstring / 清理噪音注释 |

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 34)May include surrounding context.

md
- 注释与 docstring / 噪音注释清理 / TODO 纪律 → `kb/code-comment.md`

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 18)May include surrounding context.

md
| kb/adr-changelog.md | ADR 四段结构与 supersede 流程、KaC 1.1.0 官方六类目、三模板族、release notes 区分(AC-1~11) | 写 ADR / changelog / release notes |

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 35)May include surrounding context.

md
- ADR / changelog / release notes / 决策记录流程 → `kb/adr-changelog.md`

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 21)May include surrounding context.

md
| kb/doc-governance.md | 文档漂移检测、SSOT 设计、文档门禁 CI、生命周期处置、owner 责任模型、度量局限(DG-1~11) | 治理文档腐化 / 建门禁 |

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 38)May include surrounding context.

md
- 文档漂移与腐化 / SSOT / 文档门禁 CI / owner 制 → `kb/doc-governance.md`

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 24)May include surrounding context.

md
| kb/proj-docs.md | 软件开发项目专业文档:SRS 需求规格、架构总览(C4 模型)、业务流程(BPMN)、交互设计交接、数据设计(数据字典/ER/迁移)、集成契约、项目管理(章程/纪要/状态报告/风险登记册/RACI)、安全文档(威胁建模/渗透报告)——法则 PJ-1~15、十六套可复制模板正文 |

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 41)May include surrounding context.

md
- SRS 需求规格 / 架构总览(C4)/ 业务流程(BPMN)/ 交互交接 / 数据字典·ER·迁移 / 集成契约 / 章程·纪要·状态报告·风险登记册·RACI / 威胁建模·渗透报告 → `kb/proj-docs.md`

The doc-test workflow requires executing documentation examples without warning about their potential side effects.

Medium
Category
Output Handling
Confidence
91% confidence
Finding

The workflow calls for extracting and executing documentation examples in CI, including AI-generated content, without defining a trust boundary or execution isolation. A malicious or unsafe example could access available CI credentials, modify files, or make network requests; executable-block labels and output assertions do not prevent these side effects. Human factual review elsewhere in the document does not explicitly gate execution.

Content

No source excerpt is available for this finding.

The feedback workflow instructs integrating reader behavioral data without a privacy warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

L167–L168 direct the integration and analysis of search logs, page dwell/bounce behavior, and task-failure signals. Search terms and behavioral records can contain sensitive or identifying information, but the workflow provides no privacy notice or guidance on appropriate handling.

Content

No source excerpt is available for this finding.