brainstorm-to-brief
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches generated images from well-known AI services including OpenAI, Google, and fal.ai. These are recognized technology providers, and the interactions are consistent with the skill's primary purpose of design exploration.
- [COMMAND_EXECUTION]: The skill uses local shell commands such as
mkdirandcpto organize project directories and copy internal asset templates to the project workspace. These operations are limited to the user's local project environment. - [REMOTE_CODE_EXECUTION]: The skill includes Python code snippets that utilize the
openailibrary to generate images. These scripts execute locally to interact with official remote APIs, which is the expected behavior for image generation tasks. - [DATA_EXPOSURE]: The skill generates Markdown and HTML documentation containing project details provided by the user. This data is stored locally within the user's project directory and is not exfiltrated to unauthorized third parties.
Audit Metadata