Magic Internet Access
ReviewAudited by ClawScan on May 3, 2026.
Overview
The skill appears to do what it claims—find and test free proxy nodes—but those nodes are public and untrusted, so users should use them cautiously.
Before installing, understand that this skill contacts public GitHub sources, tests third-party proxy endpoints, and gives you proxy configs from untrusted public nodes. It does not show credential theft or hidden exfiltration in the provided code, but free proxies are risky: do not use them for banking, email, work accounts, or other sensitive activity.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If a user imports these proxy nodes, their traffic may pass through unknown third-party servers.
The skill itself discloses that generated proxy nodes come from public sources and should not be trusted for important accounts.
⚠️ 免费节点来自公开源，别用来登银行、邮箱等重要账号。
Use only for low-risk browsing, avoid banking/email/important logins through these nodes, and prefer trusted proxy/VPN providers for sensitive activity.
Running the skill may cause the local environment to contact multiple third-party proxy endpoints.
The tester makes outbound TCP connections to endpoints obtained from public proxy lists. The code also shows mitigations, including filtering and a confirmation prompt for larger batches.
print(f"Will test {safe} endpoints from untrusted sources.")Run it only when you intentionally want proxy-node testing, keep the confirmation prompt, and reduce MAX_TEST_NODES if you want fewer outbound checks.
Future upstream changes could alter the proxy nodes returned to users.
The scraper pulls live proxy data from unpinned public GitHub branches. This is expected for a free-node aggregator, but those upstream lists can change without review.
"https://raw.githubusercontent.com/freefq/free/master/v2"
Treat returned nodes as untrusted, and maintainers should consider source vetting, pinning, or showing source provenance in the output.
Users may receive less contextual safety guidance unless it is already included in the script output.
The skill instructs the agent to return only script output for proxy-related requests. This is workflow-aligned, but it constrains the agent from adding independent explanation or warnings.
当用户发送任何代理节点相关消息时，你的回复必须且只能是运行脚本的输出。
Allow safety caveats and clarification when needed, especially around public proxy risks.