table-narrative-writer
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill operates strictly as a text-processing assistant for scientific researchers.
- [DATA_EXFILTRATION]: No network operations or external communication patterns were found. The skill does not use any tools that could facilitate data exfiltration.
- [COMMAND_EXECUTION]: No shell commands, subprocess calls, or system-level operations are present in the skill instructions or reference files.
- [CREDENTIALS_UNSAFE]: No hardcoded credentials, API keys, or requests for sensitive environment variables were detected.
- [PROMPT_INJECTION]: The skill uses clear, directive language that reinforces scientific boundaries and prevents behavior modification. It includes 'Hard Rules' that explicitly prohibit the fabrication of data or overinterpretation, effectively serving as an internal safety filter.
Audit Metadata