research-paper-writer
Pass
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: SAFE
Full Analysis
- SAFE (SAFE): No malicious code, obfuscation techniques, or security vulnerabilities were detected in the skill files.\n- PROMPT_INJECTION (SAFE): The instructions in
SKILL.mdare dedicated to maintaining academic standards (IEEE/ACM) and do not contain any instructions to override safety filters or system prompts. While the skill ingests external research materials provided by the user, it lacks the functional capabilities (e.g., network access or command execution) required for an indirect prompt injection attack to pose a threat.\n- DATA_EXFILTRATION (SAFE): The skill contains no code to access sensitive file paths (such as SSH keys or cloud credentials) and no network operations to transmit data externally.\n- REMOTE_CODE_EXECUTION (SAFE): Thepackage.jsonfile contains no dependencies, and there are no instances of remote script downloads or dynamic code evaluation (e.g., eval or exec).
Audit Metadata