The Agent Skills Directory

[SAFE]: The skill provides instructions for authenticating with the dreamina-canvas CLI using a standard device-authorization flow. It explicitly prohibits the storage, logging, or echoing of sensitive credentials like OAuth tokens, cookies, or session identifiers, which protects against data exposure.
[SAFE]: The workflow mandates server-side identity verification via the auth account command to ensure the session is valid before proceeding, rather than relying on potentially stale local configuration files.
[SAFE]: The skill implements strict profile isolation rules, ensuring that authentication states and canvas contexts remain segregated across different user environments (e.g., personal vs. professional) via the --profile flag.
[SAFE]: The documentation includes detailed error recovery procedures and security 'red lines,' instructing the agent to never auto-retry on permission denials or attempt to bypass browser-based authorization prompts, maintaining user control over the authentication process.
[SAFE]: The use of npx skills add to install additional components from the same vendor (full-aigc-skills) constitutes standard package management within the skill's ecosystem and does not involve untrusted or suspicious sources.

dreamina-canvas-cli-auth