implement-issues

SUSPICIOUS: the core GitHub automation is coherent with the stated purpose and explicit tool usage (`gh`, `git`) is benign, but the skill has moderate risk from transitive dependency on unspecified other skills and from processing untrusted issue/review text while dispatched agents can edit code and execute checks. No direct credential theft, exfiltration endpoint, malicious installer, or incompatible capability is evident.