constraint-input-router
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No instructions designed to bypass safety filters or override agent constraints were found. The skill emphasizes technical accuracy and the verification of visual evidence over linguistic intent.\n- [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or suspicious network operations were detected. The skill operates within a local creative workflow context.\n- [REMOTE_CODE_EXECUTION]: There are no patterns for downloading or executing external scripts or installing unverified packages. Transitions to other skills are handled through internal routing logic.\n- [OBFUSCATION]: The skill content is clear and readable. No hidden payloads, homoglyphs, or encoded URLs (such as Base64 or hex) were identified.\n- [DYNAMIC_CONTEXT_INJECTION]: The skill does not use shell command execution placeholders or other dynamic injection techniques in its documentation or configuration.
Audit Metadata