kai-topical-map
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill logic is focused on legitimate content strategy and SEO research using industry-standard methodologies.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface (Phase 0) where it reads project files like
package.jsonandREADME.mdto extract metadata for product discovery. While this represents a technical surface for indirect injection, it is a primary function for its intended use case and shows no evidence of malicious exploitation or capability abuse. - [SAFE]: Benchmark sources and external URLs point to reputable marketing technology platforms and official patent documentation.
- [SAFE]: No evidence of obfuscation, unauthorized persistence, or privilege escalation attempts was found in the instructions or referenced local paths.
Audit Metadata