thinking-via-negativa
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill provides purely instructional content for a cognitive framework ("Via Negativa") and does not include any executable code, shell commands, or network operations.
- [PROMPT_INJECTION]: No attempts to bypass safety filters or override system instructions were detected. The content focuses on architectural and procedural simplification.
- [DATA_EXFILTRATION]: The skill does not perform network calls or access sensitive local files. There are no patterns suggesting data collection or transmission.
- [REMOTE_CODE_EXECUTION]: No external dependencies, package installations, or remote script downloads are defined in the skill.
- [PRIVILEGE_ESCALATION]: The skill does not include commands to modify system permissions or acquire administrative rights. In fact, it explicitly warns against removing 'load-bearing' safety controls.
- [OBFUSCATION]: The content is written in clear, plain-text markdown with no hidden characters, encoded strings, or homoglyph attacks.
Audit Metadata