skillboss

SUSPICIOUS: the skill is internally coherent as a commercial multi-service gateway, but its footprint is unusually broad and routes credentials and user data through centralized HeyBoss infrastructure rather than direct vendor APIs. Same-org official docs reduce malware likelihood, yet the transitive install/update model, credential forwarding, and third-party gateway data flows make it medium-high security risk.

No clear signs of embedded runtime malware in this fragment. However, the code is a high-impact filesystem-to-network uploader that recursively reads project files and explicitly allows uploading a local `.env` file. Combined with a user-overridable apiUrl and authenticated upload of packaged contents/bindings/migrations, the primary risk is supply-chain style data exfiltration to an untrusted or attacker-controlled deployment endpoint.

No strong evidence of direct malware (no persistence, no obvious data theft, no obfuscation, no suspicious network destinations beyond the expected backend endpoints). However, this module contains a high-impact security weakness: openBrowser() uses execSync with a shell command string that interpolates a URL returned from a backend response, creating a potential command-injection pathway if the backend (or its response) can be influenced. Additionally, the module appends an x-vercel-protection-bypass parameter derived from a configured token to outbound requests, which is unusual and should be reviewed/strictly validated server-side. Overall risk is medium-to-high for environments where backend responses could be malicious or tampered; otherwise it is primarily an operational hardening concern.

No direct indicators of overt malware (e.g., backdoor/persistence, credential theft, or exfiltration) are present in this updater wrapper. However, it has a significant supply-chain integrity weakness: it downloads a remote ZIP authenticated by an API key and extracts it into the installation directory with -Force without verifying integrity/signature or constraining/validating the archive contents before installation. This means a compromised update endpoint or tampered artifact could lead to arbitrary code being installed on the client.

No direct malicious behavior (e.g., reverse shells, credential theft, or network exfiltration) is evident in the script itself. However, it is a high-trust supply-chain installer: it authenticates with a locally stored API key and then installs remote code by extracting an untrusted zip archive without cryptographic integrity/authenticity verification. It also extracts without explicit path/symlink safeguards, leaving potential exposure to archive path traversal/overwrite risks depending on unzip behavior. The primary risk is artifact-compromise impact (server/account compromise or tampering) rather than malware embedded directly in this updater script.

This fragment is static JSON configuration with embedded external service endpoints and a plaintext API key. While it contains no executable malicious logic by itself, the hardcoded secret is a critical supply-chain security weakness: the key is likely to be used for authenticated outbound requests and should be rotated and removed from any distributed artifacts, replaced with runtime secret injection (environment variables/secret manager).