fix-issues

SUSPICIOUS: the skill is purpose-aligned and uses official GitHub paths, so it is not malware-like, but it grants an AI agent high-impact autonomous repo actions and mixes untrusted external issue content with code execution and write access. The main risk is unsafe automation, not covert exfiltration or deceptive install behavior.