This reference explains how the Vibgrate CLI behaves in different environments and how environment-level settings interact with the CLI. It is intended for developers configuring CI runners, containers, or shared workstations.
Credentials and login state
The recommended way to authenticate is the browser-based login flow:
vg login
This stores a login credential locally so subsequent commands like vg push work without copying a token. To clear it:
vg logout
For non-interactive environments such as CI, use a DSN token created with vg dsn create and supply it to the upload step. See the dashboard upload guide for details.
Network and proxy environments
If your network routes outbound HTTPS through a proxy, the CLI honors your shell's standard HTTPS proxy configuration. In restricted networks you can force fully local behavior:
vg scan --offline
Offline mode uses only local artifacts and caches and does not attempt network calls. See the proxy and offline configuration article for a deeper walkthrough.
Working directory context
When the CLI cannot rely on the shell's current directory, set it explicitly:
vg scan --cwd /path/to/project
This is useful in CI where the checkout directory differs from the runner's default location.
Caching
The CLI maintains local caches to speed up repeated work, including the semantic index used by code-map queries. To bypass the cache for a run:
vg scan --no-cache
Notes
Keep secrets out of committed config. Use your CI provider's secret store for DSN tokens and reference them only in the upload step. The CLI never needs your source code to leave the machine for a local scan.
Related
- Proxy and offline configuration
- The
vg login,vg logout, andvg dsn createcommand pages - Vibgrate Cloud upload guide