Back to skill

Security audit

xiaohongshu-post

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because scanners report actions that may disrupt unrelated browser sessions and do not reliably confirm a post is publicly visible; two scripts were only partly inspected.

The verdict is suspicious, so review the concerns before installing. In particular, consider whether you are comfortable with a setup step that may close unrelated browser sessions, verification that may not confirm the actual post, and the global CLI installation instruction. The scan was partial, so it does not establish that the scripts are safe.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (9)

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 273)May include surrounding context.

md
| `scripts/publish_note.js` | 填标题/正文/话题/传图,**默认不点发布**(`--publish` 才点) |

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 274)May include surrounding context.

md
| `scripts/verify_note.js` | 取 id/token/短链 + 无痕上下文验证公开可见 |

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Browser bootstrap invokes an unscoped playwright-cli kill-all command, potentially terminating unrelated automation sessions.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

The bootstrap executes playwright-cli kill-all rather than limiting cleanup to the skill's own session. This can terminate unrelated automation sessions and interrupt work or lose unsaved browser state, exceeding what is necessary to attach the publishing session.

Content

No source excerpt is available for this finding.

Existing playwright-cli sessions are terminated without user disclosure.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The script silently terminates existing playwright-cli sessions while presenting itself as browser bootstrap and attachment. This is the same underlying issue as the unscoped kill-all finding: unrelated user activity can be disrupted without an opportunity to approve or preserve it.

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 · scripts/verify_note.js (reported line 76)May include surrounding context.

js
// 2) 短链(需 www 域有登录态;note_id 必须用 API 给的 id,不能从当前 URL 猜)
  try {
    const key = await page.evaluate(async (nid) => {
      const r = await fetch("https://edith.xiaohongshu.com/api/sns/web/share/code", {
        method: "POST", credentials: "include",
        headers: { "content-type": "application/json" },
        body: JSON.stringify({ note_id: nid })

Public-visibility verification does not enforce the documented title and body checks.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

Without --expect, any final URL lacking /404 is accepted as publicly visible, including login, challenge, generic error, or unrendered pages. Even with --expect, the required title check is absent, allowing the agent to report successful public publication without verifying the intended content.

Content

No source excerpt is available for this finding.

The troubleshooting instructions prescribe a global npm installation without warning that it modifies the user's shared CLI environment.

Low
Category
Output Handling
Confidence
88% confidence
Finding

The remedy explicitly instructs running npm install -g @playwright/cli@latest, rather than merely listing an available tool. It does not disclose that this installs third-party software globally and may change the CLI version used by other projects.

Content

No source excerpt is available for this finding.

The dependency-loading fallback executes a shell command without disclosing that subprocess execution.

Low
Category
Output Handling
Confidence
93% confidence
Finding

When local Playwright imports fail, the script executes npm root -g through a shell. Neither the usage docstring nor a comment or user-facing message explains this subprocess operation.

Content

No source excerpt is available for this finding.