stats

The skill's stated purpose (retrieving storage usage stats) aligns with its capabilities: it reads an API key and calls the service's /stats endpoint and formats the response. There are no direct signs of code execution or obfuscation and no download-and-execute patterns. However, the skill reads a sensitive environment variable and will transmit that key to an external service, and it also auto-fetches a test key from a remote URL when the environment variable is missing. These behaviors are legitimate for an API-backed stats feature but carry moderate supply-chain/credential risk because credentials are read and forwarded without an explicit confirmation step and because auto-generation fetches a key from the network. I assess low probability of active malware (no backdoors or remote exec patterns), but a medium security risk primarily from credential exposure/forwarding and the automatic network-based key provisioning. Recommend: require explicit user consent before reading/transmitting existing env credentials, document the test-key generation behavior clearly, and allow users to opt-out or paste a key interactively instead of auto-fetching.