Back to skill

Security audit

Maimai Connector

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because scanners report instructions to run browser commands without sandbox isolation and a persistent background heartbeat, despite finding its stated purpose coherent.

Review the sandbox instructions and persistent heartbeat before installing. Consider whether you are comfortable using the skill with an authenticated browser session under reduced isolation; the scanners report concerns but do not establish malicious intent.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · README.md (reported line 9)May include surrounding context.

md
## 原理

- 认证靠**浏览器登录态**(扫码登录后下发的 session cookie),不在 header 传 token
  (用 curl 直接打接口会返回 `20002 Invalid access token`)
- 所有 `/api/ent/*` 接口必须带 `uid=<当前账号 uid>`,否则返回 `21001 Incorrect parameters`

## 安装

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 57)May include surrounding context.

md
1. **所有 `/api/ent/*` 必须带 `uid=<UID>`**,否则 `21001 Incorrect parameters`。
   UID / CO_ID 占位符硬编码在三个 js 里,使用前替换成自己的账号信息(见 README「安装」)。
2. **认证靠浏览器登录态**(扫码后下发的 session cookie),不在 header 传 token。
   解密 cookie 用 curl 打接口行不通(返回 `20002 Invalid access token`)。
3. `user/current` 里的 `people_search: false` 是误导,真实权限看
   `/api/ent/search/user_search_pre_check` 的 `in_cid_whitelist`。
4. 二维码内容 = `taoumaimai://qr_login?login_id=<qr_code>`。

The workflow recommends running all browser commands outside the sandbox without warning about the loss of isolation.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The README broadly recommends running browser commands outside the execution sandbox to resolve Chrome discovery failures, removing an isolation boundary rather than addressing the compatibility issue narrowly. Because the workflow executes adapters against an authenticated browser session, compromised adapter code or dependencies would have greater access to host resources and sensitive account data; the operational rationale does not make this safe.

Content

No source excerpt is available for this finding.

The skill requires disabling sandbox isolation without warning about the security implications.

Medium
Category
Output Handling
Confidence
99% confidence
Finding

The skill explicitly requires disabling execution sandboxing for every browser invocation, turning a browser-discovery workaround into a blanket security-boundary bypass. Authenticated browser access and the persistent background heartbeat increase the exposure if adapter code or dependencies are compromised, although the supplied text does not establish malicious intent or an actual compromise.

Content

No source excerpt is available for this finding.