cold-email
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill consists exclusively of markdown instructions for the LLM. It does not include any scripts, binaries, or command-line operations.
- [DATA_EXFILTRATION] (SAFE): There are no network-capable commands (like curl or wget) or references to sensitive local files that could lead to data leakage.
- [REMOTE_CODE_EXECUTION] (SAFE): No external packages or remote scripts are downloaded or executed. The skill does not use any dynamic execution functions.
- [PROMPT_INJECTION] (SAFE): The instructions are focused on copywriting frameworks. There are no attempts to override system prompts or bypass safety guidelines.
- [INDIRECT_PROMPT_INJECTION] (SAFE): While the skill ingests user input via $ARGUMENTS, it has no tool-use capabilities (e.g., shell access, file writing). Therefore, any malicious instructions embedded in the user input would only affect the text output of the email and cannot compromise the system.
Audit Metadata