reference-finder

[Skill Scanner] Installation of third-party script detected All findings: [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] Benign overall with moderate caution. The skill’s footprint aligns with its stated purpose (domain extraction + literature generation + Markdown output). The main concerns are best-practice configuration (API keys should be supplied via environment, not hardcoded) and proxy handling (misconfiguration could route data through untrusted proxies). No explicit credential harvesting or data exfiltration patterns are evident. The inclusion of a local proxy option is not inherently malicious but requires proper access controls and secure configuration to avoid misrouting of API calls. Overall, the design is coherent and proportional to its purpose, with recommended clarifications to credential management and proxy usage to reduce risk. LLM verification: The skill’s intended workflow is coherent and functional for automated literature reference generation. Security concerns are moderate due to documentation and configuration risks: unpinned dependencies, potential hardcoded API keys, and localhost proxy usage. To move toward a secure-by-design deployment, pin dependencies, enforce secret management (never log or commit API keys), constrain proxy usage, and remove insecure installation guidance from docs. With these mitigations, the overall risk