Email Classifier
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill file contains only markdown text and instructional metadata. There are no scripts, binaries, or configuration files that execute logic.
- [NO_CODE]: Analysis confirmed the absence of any Python, Node.js, or shell script components. All functionality is handled through the agent's natural language processing.
- [PROMPT_INJECTION]: The skill instructions define how to process external email data. However, since the skill lacks the ability to perform network requests, write to the filesystem, or execute commands, there is no viable path for an indirect prompt injection attack to cause system-level harm.
Audit Metadata