saccoai-client-portal
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
opencommand to present the generated static review portal to the user. It also utilizes thevercelCLI to manage deployments and cloud marketplace resources. - [EXTERNAL_DOWNLOADS]: The skill interacts with Vercel's infrastructure to deploy Next.js applications and provision Upstash Redis instances for real-time feedback storage.
- [PROMPT_INJECTION]: The skill processes untrusted input from client-provided feedback which is used to inform downstream cleanup and merge tasks.
- Ingestion points: Reads user-provided
feedback.jsonfiles and fetches feedback strings from remote Redis databases. - Boundary markers: None identified; the skill directly parses and presents the content of the feedback fields.
- Capability inventory: Includes shell command execution (
open,vercel) and project cleanup capabilities via integration with thesaccoai-design-variationsskill. - Sanitization: No explicit sanitization or filtering of the feedback text is performed before it is presented to the user or passed to the cleanup phase.
Audit Metadata