analise-semiotica
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a specialized persona for semiotic analysis and does not exhibit any malicious behaviors.
- [PROMPT_INJECTION]: No attempts to override system instructions or bypass safety filters were found. The instructions focus on structured analysis and educational feedback.
- [DATA_EXFILTRATION]: There is no access to sensitive local files, environment variables, or hardcoded credentials. Network activity is limited to legitimate use of a search tool for finding educational resources.
- [REMOTE_CODE_EXECUTION]: The skill does not include any functionality to download, install, or execute external scripts or packages.
- [COMMAND_EXECUTION]: No dangerous system commands, privilege escalation attempts, or persistence mechanisms were detected.
- [INDIRECT_PROMPT_INJECTION]: 1. Ingestion points: User-provided images/videos and justification text in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Uses google_search; no subprocess, file-write, or network calls beyond tool use. 4. Sanitization: No explicit sanitization of external content is defined. The risk is low as capabilities are restricted to non-destructive actions.
Audit Metadata