sync-melech-skills

SUSPICIOUS: the skill is internally consistent as a global skill-sync meta-tool, but its actual footprint is broad and high-impact. The main concern is transitive installation of many downstream skills into every agent with non-interactive global changes, creating a substantial trust-chain risk even without clear evidence of direct credential theft or exfiltration.

No direct malware primitives are evident in this fragment. The main security concern is supply-chain and automation risk: it executes local status.py and then uses `npm exec --package=skills` to run a CLI that can add/update skills, with arguments largely driven by status.py JSON (including skill names) and with a wildcard scope ('-a', '*'). If the npm package, the canonical source resolution, or status.py outputs/inputs are compromised, the resulting external command actions could be harmful. Recommend pinning/verifying the npm package/CLI version, validating catalog fields before passing them to subprocess, and reviewing status.py and the invoked npm/CLI behavior for integrity checks and safety controls.