recon-github-code-leaks

SUSPICIOUS. The skill's capabilities are internally consistent with its stated purpose, but the purpose itself is offensive secret-hunting and foothold discovery against third-party organizations. Data flows mostly stay on official platforms and AWS, with no clear credential harvesting proxy or covert exfiltration, so this is not confirmed malware; however, enabling an AI agent to scan public code history for secrets and test found credentials creates high misuse risk.