Lingxi

This package's docs describe a powerful orchestration system that needs secrets, installs packages, writes token files, and starts a web dashboard, but the registry metadata doesn't list any of those requirements. Before installing or running it: - Treat the code as untrusted until you can verify the referenced scripts actually exist and match the docs (the manifest here contains only docs, no code files mentioned in SKILL.md). - Do not place real credentials (GitHub token, cloud API keys, platform bot tokens) into the paths the README suggests until you confirm the code that will read them and you trust its origin. - Ask the publisher to provide accurate metadata: a clear install spec, exact required env vars/config paths, and a signed repository or official release URL. - Prefer testing in an isolated sandbox/VM with no access to sensitive networks/secrets; if you must run on a host, use a throwaway account and no production credentials. - Verify the GitHub repository (https://github.com/AI-Scarlett/lingxi-ai) and the presence and integrity of the scripts (orchestrator, executors, dashboard) before giving the skill filesystem/network access. Given the mismatches and the instructions that would write persistent tokens and open network services, proceed cautiously; the package is suspicious until these inconsistencies are clarified.