The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands like grep and printf to manage local configuration files within the ~/.cyrus/ directory. It also utilizes the vendor-specific CLI tool cyrus to perform OAuth authorization. These actions are limited to the skill's primary purpose and do not represent arbitrary execution risks.
[DATA_EXPOSURE]: The skill accesses configuration files (~/.cyrus/.env and ~/.cyrus/config.json) to verify setup status. It incorporates strict constraints to prevent the agent from reading secret values into the chat context, instead using pattern-matching commands to verify the existence of keys without exposing their content.
[EXTERNAL_DOWNLOADS]: The skill utilizes browser automation (via agent-browser or MCP tools) to navigate to the official Linear settings page (linear.app) to facilitate OAuth application creation. This interaction is transparent and aligned with the skill's core functionality.
[CREDENTIALS_UNSAFE]: While the skill manages credentials, it explicitly instructs the user to manually copy and paste secrets into the environment file. It includes a critical safety guideline forbidding the agent from scraping or extracting secrets from the browser or reading them from the filesystem.

cyrus-setup-linear