@fxri/toolkit@1.10.1
专为多人 + AI 跨项目协作打造:任务管理 + 多语言 CHANGELOG。
AI security review
scanned · by lpm-firewall-aiThe source confirms automatic agent skill directory changes during normal CLI startup, a guarded package-owned extension setup risk. The inspected network path only checks the package version and does not substantiate command-output exfiltration, so the evidence supports a warning rather than a block.