Skip to main content

End user with Aether CLI guide

This guide is for users whose main experience is the Aether CLI and a local mount.

You do not need to learn every HTTP route to be effective in this role.

What matters most in this role​

As an Aether CLI end user, the highest-signal concepts are:

  • remote endpoint
  • mount workflow
  • local cache
  • logging and metrics
  • when to use the mounted workspace versus the product UI

Your core workflow​

The normal Aether CLI end-user flow is:

  1. get a session ID
  2. configure the endpoint
  3. mount the session locally
  4. work in your editor or shell
  5. use the product or API-driven workflow for review, approval, checkpoint, or persistence when required
  6. stop or unmount when finished

Read these pages first​

Start here:

Then use:

What you should configure explicitly​

At minimum:

  • service endpoint
  • cache root
  • log format if support or automation needs it

If you work across several environments:

  • separate cache roots by environment
  • use explicit wrappers or shell profiles

What belongs in Aether CLI versus the product UI​

Use Aether CLI for:

  • local filesystem work
  • editors
  • shell tools
  • interactive changes on your machine

Use the product or API-driven workflow for:

  • review metadata
  • approvals
  • source creation and promotion
  • reporting and account-level visibility

How to think about performance​

When a mount feels wrong, the likely user-controlled causes are:

  • bad endpoint configuration
  • stale cache assumptions
  • TTL settings that do not match the workflow
  • excessive or missing diagnostic logging when support is needed

Common Aether CLI end-user mistakes​

Avoid:

  • mixing environments in one cache root
  • expecting aether status to behave like a generic connectivity probe
  • using Aether CLI as if it were the server
  • putting secrets into the mounted workspace by default

Fastest useful Aether CLI toolkit​

If you only remember a short path:

  1. endpoint
  2. mount
  3. status and doctor
  4. cache/logging/metrics only when needed

That is enough for most end users to be productive without learning the full product internals.