Skill Trust Decision

game-ui-voice-pack

Game UI Voice Pack 是合法的语音合成工具包,代码质量高,实现了完整的 OAuth 认证、MCP 通信和自动更新机制。主要风险点为自动更新功能的供应链理论风险,但代码包含 SHA256 验证可有效缓解。

Install decision first Source: ClawHub Scanned: Aug 24, 2026
Files 13
Artifacts 7
Violations 0
Findings 4

Why this conclusion was reached

0/4 dimensions flagged
Pass
Declared vs actual capability

Declared resources and inferred behavior are broadly aligned.

Review
Hidden execution and egress

7 lower-risk artifacts were extracted and still need context.

Pass
Attack chain and severe findings

There is no explicit malicious chain in the report.

Pass
Dependencies and supply chain hygiene

Dependencies are present but no obvious high-risk issue stands out.

What drove the risk score up

自动更新供应链风险 +10

代码从 cdn.beatra.ai 下载并自动替换 package-owned 文件,虽有 SHA256 验证但缺少代码签名

未声明的系统信息收集 +5

detect_host_platform() 和 _record_skill_inventory() 收集 agent 类型/平台信息,SKILL.md 未提及

权限声明宽泛 +5

allowed-tools 未明确声明 filesystem:WRITE 和 network:WRITE 权限

Most important evidence

Low Supply Chain

自动更新功能存在供应链理论风险

代码实现自动更新机制,可从 cdn.beatra.ai 下载并自动替换 package-owned 文件(包括 scripts/*.py)。虽包含 SHA256 验证(discovery、manifest、archive 三级验证),但缺少代码签名验证。攻击者若能入侵 CDN 可能注入恶意代码。

scripts/mcp_client.py:31
建议添加代码签名验证作为额外安全层,当前 SHA256 多级验证已提供较强保护。
Low Doc Mismatch

未声明的系统信息收集

detect_host_platform() 函数读取 AI_AGENT、CLAUDECODE、CODEX_* 等环境变量检测 agent 平台;_record_skill_inventory() 记录安装路径。SKILL.md 未提及这些数据收集行为。

scripts/authorize.py:343
在 SKILL.md 的隐私/数据收集部分添加说明,告知收集 agent 类型和安装路径用于遥测。
Info Priv Escalation

credential 文件权限验证

代码在 POSIX 系统上验证 ~/.beatra 目录权限为 700,credentials.json 为 600,确保凭证文件仅对当前用户可读。这是良好的安全实践。

scripts/mcp_client.py:1022
无建议,当前实现符合最佳实践。
Info Supply Chain

事务性更新和回滚机制完善

自动更新实现完整的事务机制:写入临时文件、创建备份、journal 记录、回滚函数。即使更新中断也能恢复到一致状态。

scripts/mcp_client.py:520
无建议,当前实现质量高。

Declared capability vs actual capability

Filesystem Pass
Declared NONE
→
Inferred WRITE
scripts/authorize.py:206 创建 ~/.beatra/ 目录,mcp_client.py:723 写入配置文件
Network Pass
Declared READ
→
Inferred WRITE
scripts/authorize.py:25 向 api.beatra.ai 发送 OAuth 请求,scripts/mcp_client.py:29 调用 MCP 端点
Environment Pass
Declared NONE
→
Inferred READ
scripts/authorize.py:343 读取 AI_AGENT、CLAUDECODE 等环境变量检测平台

Suspicious artifacts and egress

Medium External URL
https://console.beatra.ai/wallet?intent=buy

SKILL.md:88

Medium External URL
https://beatra.ai/skills/game-ui-voice-pack

manifest.json:5

Medium External URL
https://beatra.ai/icon-512x512.png

manifest.json:6

Medium External URL
https://mcp.beatra.ai/mcp

manifest.json:56

Medium External URL
https://api.beatra.ai

scripts/authorize.py:25

Medium External URL
https://beatra.ai/skills/game-ui-voice-pack/channels/clawhub/install.json

scripts/mcp_client.py:31

Medium External URL
https://cdn.beatra.ai/agent-packages/game-ui-voice-pack/channels/clawhub/v

scripts/mcp_client.py:32

Dependencies and supply chain

PackageVersionSourceKnown vulnNotes
urllib stdlib Python 3 标准库 No 使用标准库进行 HTTP 请求,无第三方依赖

File composition

13 files · 3161 lines
Python 3 files · 2490 linesMarkdown 9 files · 610 linesJSON 1 files · 61 lines
Files of concern · 4
scripts/mcp_client.py Python · 1565 lines
自动更新功能存在供应链理论风险 · credential 文件权限验证 · 事务性更新和回滚机制完善 · https://beatra.ai/skills/game-ui-voice-pack/channels/clawhub/install.json · https://cdn.beatra.ai/agent-packages/game-ui-voice-pack/channels/clawhub/v
scripts/authorize.py Python · 619 lines
未声明的系统信息收集 · https://api.beatra.ai
SKILL.md Markdown · 189 lines
https://console.beatra.ai/wallet?intent=buy
manifest.json JSON · 61 lines
https://beatra.ai/skills/game-ui-voice-pack · https://beatra.ai/icon-512x512.png · https://mcp.beatra.ai/mcp
Other files · uninstall.py · installation-and-auth.md · billing-errors-and-recovery.md · uninstall-and-disconnect.md · workflow.md · tasks-and-results.md +2

Security positives

代码质量高,包含完整的错误处理和类型注解
OAuth 2.0 Device Authorization 流程标准且安全
凭证文件权限验证严格(目录 700/文件 600)
更新机制包含 SHA256 三级验证(discovery/manifest/archive)
事务性更新实现完整,支持原子性回滚
无恶意指标(无 base64 编码执行、无远程 shell、无凭证外传)
所有外部 URL 均为官方域名(beatra.ai、cdn.beatra.ai)
MCP 协议实现规范,包含 session 管理和错误处理