Back to skill

Security audit

内容裂变工厂

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly coherent for content automation, but it can automatically create and share Feishu documents from AI-generated public-trend content without a clear approval gate.

Install only if you are comfortable with the skill contacting public content sources, generating articles/cards, and creating Feishu documents by default. Prefer running with draft mode first, review generated content and links before sharing, restrict the Feishu bot to a dedicated folder, and avoid scheduled runs until approval and validation steps are added.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
workflow.json:57
Finding

Untrusted Public Content Can Influence Default-On Feishu Publication

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger phrases are broad natural-language requests like '生成今日内容选题' and '热点文章一键搞定', which increases the chance of accidental invocation during ordinary conversation. Because this skill performs a multi-step workflow culminating in external writes to Feishu, overbroad activation can cause unintended data collection, content generation, and document publication.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill automatically creates and writes a Feishu document in Step 5, but the description and usage flow do not clearly warn users up front that running the skill causes an external side effect. This is dangerous because users may believe they are only generating drafts or analysis locally, while the workflow actually publishes content to a connected platform unless they know to use a draft mode.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill's security note claims Feishu documents are private by default, but the workflow explicitly sets grant_to_requester: true, which weakens the stated privacy posture and can disclose generated content to whoever triggers the skill. In an agent context, implicit document sharing is risky because the output may include scraped trend data, internal analysis, or drafted content that the operator did not intend to expose automatically.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The workflow is framed as a full-chain automation factory for trending discovery, content generation, card rendering, and publishing, but it does not declare meaningful activation boundaries, approval gates, or exclusions. That broad scope increases the chance a user triggers web collection and downstream publication behaviors without understanding the operational and reputational consequences.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The workflow automatically creates a Feishu document containing AI-generated content and grants access to the requester when draft mode is false, yet the manifest does not present a prominent warning or explicit consent checkpoint for this external write action. This can lead to unintended data disclosure, reputational harm, or propagation of inaccurate/generated content into a shared organizational workspace.

Content

No source excerpt is available for this finding.