扫描报告
5 /100
webmcp-bridge
Connect a website to the local-mcp browser bridge through a fixed UXC link. Use when the user needs to operate native WebMCP sites or adapter-backed sites through local-mcp, manage per-site browser profiles, or switch bridge presentation modes explicitly.
The WebMCP Bridge skill is a legitimate browser automation tool that creates CLI links for managing browser sessions through local-mcp, with well-documented shell usage and no malicious behavior detected.
可以安装
This skill can be used safely. The shell execution and npx usage are necessary and documented. Consider pinning the @webmcp-bridge/local-mcp version in production environments for reproducibility.
安全发现 1 项
| 严重性 | 安全发现 | 位置 |
|---|---|---|
| 低危 | Unpinned npm package dependency 供应链 | scripts/ensure-links.sh:28 |
| 资源类型 | 声明权限 | 推断权限 | 状态 | 证据 |
|---|---|---|---|---|
| 命令执行 | WRITE | WRITE | ✓ 一致 | SKILL.md:scripts/ensure-links.sh |
| 文件系统 | READ | WRITE | ✓ 一致 | Creates ~/.uxc/webmcp-profile/<site> directories for browser profiles |
| 网络访问 | READ | READ | ✓ 一致 | Browser automation fetches target URLs; external URLs documented in references |
| 浏览器 | READ | WRITE | ✓ 一致 | Manages browser profiles, sessions, and automation through local-mcp bridge |
2 项发现
中危 外部 URL 外部 URL
https://board.holon.run references/source-modes.md:12 中危 外部 URL 外部 URL
http://127.0.0.1:4173 references/source-modes.md:13 目录结构
8 文件 · 19.2 KB · 657 行 Markdown 5f · 462L
Shell 2f · 188L
YAML 1f · 7L
├─
▾
agents
│ └─
openai.yaml
YAML
├─
▾
references
│ ├─
link-patterns.md
Markdown
│ ├─
source-modes.md
Markdown
│ ├─
troubleshooting.md
Markdown
│ └─
usage-patterns.md
Markdown
├─
▾
scripts
│ ├─
ensure-links.sh
Shell
│ └─
validate.sh
Shell
└─
SKILL.md
Markdown
依赖分析 3 项
| 包名 | 版本 | 来源 | 已知漏洞 | 备注 |
|---|---|---|---|---|
@webmcp-bridge/local-mcp | unpinned | npm | 否 | No version pinning - uses latest with npx -y |
uxc | unpinned | system | 否 | Expected to be pre-installed by user |
npx | unpinned | system | 否 | Expected to be pre-installed by user |
安全亮点
✓ Shell usage is fully documented in SKILL.md with clear explanation of each command
✓ No credential harvesting or sensitive data exfiltration detected
✓ No obfuscation, base64 encoding, or anti-analysis techniques observed
✓ Clear separation of concerns with well-structured reference documentation
✓ No unauthorized resource access to ~/.ssh, ~/.aws, .env or similar sensitive paths
✓ Guardrails are explicitly documented in SKILL.md (profile isolation, no credential script extraction)
✓ Explicit profile isolation guidance prevents cross-site data leakage
✓ Command validation script (validate.sh) helps catch documentation drift