deep-research
Fail
Audited by Socket on Mar 10, 2026
1 alert found:
Obfuscated FileObfuscated FileSKILL.md
HIGHObfuscated FileHIGH
SKILL.md
The Deep Research Skill aligns with its stated purpose of autonomous, multi-step research using the Gemini API and producing structured reports. The main security considerations are standard credential handling for an API key and ensuring outputs do not inadvertently leak credentials or internal logs. No evident malicious data flows or undesired network exfiltration paths are described. The footprint seems coherent and proportionate to the intended use, with modest security risk primarily around credential exposure and per-step user approvals for autonomous actions.
Confidence: 98%
Audit Metadata