obsidian-second-brain

SUSPICIOUS: the skill's purpose broadly matches vault management and research, but its footprint is large for a 'front door' skill: transitive skill installation, curl|bash and clone-and-run installers, persistent hooks, and autonomous background/scheduled behavior. Data flows mostly align with the stated purpose and official services, so this is not confirmed malware, but it is higher-risk than a typical documentation/router skill.

No clear indicators of intentional malware (no obfuscation, no exfiltration, no backdoor/persistence beyond an expected jeo hook, no credential theft). However, the script increases supply-chain risk by fetching and executing third-party code (npx from REPO_URL; git clone/pull from UPSTREAM_URL; executing upstream scripts; installing a runtime hook that will execute the resolved adapter). Without version pinning/integrity checks in this script, compromise of those sources could lead to arbitrary code execution in the user’s environment.