技能 maxhub-toutiao 审计历史
📰

审计历史

maxhub-toutiao - 3 审计

审计版本 3

最新 安全

May 20, 2026, 02:50 PM

All 121 static analysis findings were evaluated and dismissed as false positives. The skill contains no executable code - all files are markdown documentation and JSON metadata. The 53 external_commands findings are curl command examples inside markdown code blocks, not executable code. The 14 weak cryptographic algorithm findings are API response format strings (e.g., {code, message, data}) and example tokens in documentation. The 9 system reconnaissance findings are parameter description tables. The high entropy finding is a false positive from Chinese text. The critical dangerous combination finding is an aggregation of the above false positives. The skill is a legitimate read-only API documentation reference that requires a MAXHUB_API_KEY environment variable to query Toutiao data endpoints.

6
已扫描文件
489
分析行数
3
发现项
claude
审计者
未发现安全问题

风险因素

🌐 网络访问 (25)
⚙️ 外部命令 (53)
🔑 环境变量 (15)

审计版本 2

低风险

May 20, 2026, 01:10 PM

Static analysis flagged 121 potential issues, but all high-severity findings are false positives caused by the analyzer misreading markdown documentation as executable code. The skill is a legitimate Toutiao API client that uses curl to query data from the MaxHub API (aconfig.cn). All network URLs, shell command examples, and environment variable references are in documentation or code blocks showing legitimate usage. The expected risks (network access, env_access for MAXHUB_API_KEY, curl external commands) are standard for an API client and are clearly declared.

6
已扫描文件
489
分析行数
6
发现项
claude
审计者
中风险问题 (1)
Environment variable access for API key
The skill reads the MAXHUB_API_KEY environment variable to authenticate with the MaxHub API. This is expected and declared behavior for an API client skill. The key is used only as a Bearer token for API requests to aconfig.cn.
低风险问题 (2)
Network URL references in documentation
URLs pointing to aconfig.cn and toutiao.com appear in documentation files. These are legitimate API endpoint references and example URLs, not executable network calls. Expected for an API client skill.
Shell command examples in markdown
Backtick-style commands appear in documentation (curl, export, npx commands). All are inside markdown code blocks showing example usage. No commands are executed dynamically. False positive from static analyzer misreading markdown formatting as executable backtick commands.

审计版本 1

低风险

May 9, 2026, 07:35 AM

All 70 static findings are false positives. The detected patterns (external_commands, env_access, hardcoded URLs) are markdown documentation with curl command examples and environment variable references. The skill is legitimate news data collection for the Toutiao platform with proper security declarations, rate limits (60 calls/minute, max 5 pages), and cost controls. No malicious behavior confirmed.

3
已扫描文件
284
分析行数
3
发现项
claude
审计者
未发现安全问题

风险因素

⚙️ 外部命令
未记录任何特定位置
🌐 网络访问 (2)
🔑 环境变量 (1)