mino-driven-design-skills
Fail
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The installation guide explicitly instructs users to run PowerShell scripts with the
-ExecutionPolicy Bypassflag. This technique is used to circumvent local security configurations that prevent the execution of untrusted scripts on Windows systems. - [EXTERNAL_DOWNLOADS]: The skill directs users to clone a repository from
https://github.com/my-take-dev/inspired-mino-design-skills.git. This repository and organization are not established as trusted sources, posing a risk of downloading malicious content. - [COMMAND_EXECUTION]: Users are instructed to execute shell and PowerShell scripts (
.agents/validate-suite.shand.agents/validate-suite.ps1) immediately after cloning the repository. This facilitates the execution of arbitrary, potentially malicious code from an external source. - [PROMPT_INJECTION]: The skill's primary function is to ingest and process user-provided feature requirements and descriptions. However, it lacks instructions for the agent to implement boundary markers or delimiters when handling this untrusted data, which could allow instructions embedded within the user data to hijack the agent's behavior.
- [CREDENTIALS_UNSAFE]: While the skill correctly advises users not to commit secrets and to use environment variables for
MINO_VALIDATOR_API_KEYandEVIDENCE_UPLOAD_TOKEN, the mention of these variables indicates the skill may handle sensitive tokens which could be targeted for exfiltration if the local scripts are compromised.
Recommendations
- AI detected serious security threats
Audit Metadata