The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on shell commands using npx @globalize-now/cli-client and git to perform platform operations. A minor inconsistency was noted where the global binary is referred to as globalise-now-cli while the package and organization use globalize-now.
[EXTERNAL_DOWNLOADS]: The skill uses npx to download and run the @globalize-now/cli-client package. This package is owned by the author ('globalize-now') and is used for its intended platform management functions.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external repositories and CLI outputs, creating a surface for potential indirect prompt injection.
Ingestion points: Git remote URLs (git remote get-url), repository path structures (github detect), and contents of locale files (repositories detect).
Boundary markers: The agent is instructed to use the --json flag for all commands to ensure output is handled as structured data. A mandatory 'CONSENT GATE' (Step 2c.3) requires the user to explicitly confirm repository details before the agent connects them to the platform.
Capability inventory: The skill can execute various CLI commands to create, update, or delete resources (projects, glossaries, members, etc.) and perform git operations across multiple scripts.
Sanitization: The instructions emphasize matching specific UUIDs and numeric IDs from parsed JSON output rather than directly interpolating string content into commands.
[CREDENTIALS_UNSAFE]: The skill includes instructions for managing API keys (api-keys list/create/revoke). While these are standard management functions, they provide the agent with the ability to view and generate sensitive authentication tokens for the organization.

globalize-now-cli-use