open-design-mode

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to interact with the local open-design and codex CLI tools for tasks such as installation verification, login status checks, and executing design runs. These commands are restricted to the local environment and are part of the intended functionality for managing the design runtime.
  • [EXTERNAL_DOWNLOADS]: The instructions direct the agent to provide the user with the official download link at https://open-design.ai/download/ if the required software is missing. This is a standard and safe practice for ensuring the user obtains the correct, verified software directly from the vendor.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external artifact data and tool outputs. It proactively mitigates the risk of recursive injection or tool confusion by including explicit prompt boundaries that instruct the child-runtime not to re-invoke the Open Design plugin or login workflows.
  • [CREDENTIALS_UNSAFE]: The skill follows security best practices by explicitly instructing the agent never to request, store, or transmit raw API keys or cloud credentials in chat or plugin files. It instead utilizes secure authentication flows like activation URLs and stdin-based key entry.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 06:34 AM
Security Audit — agent-trust-hub — open-design-mode