Use Amics from a coding agent

The agent stays in the local repository where it can edit, test, compare results, and commit. Amics handles one bounded remote command at a time.

Recommended setup

1. Install the standalone amics CLI and run amics login.
2. Install the Amics agent skill.
3. Ask the agent for a read-only GPU recommendation.
4. Approve an explicit maximum hourly price, runtime, and optional warm window.
5. Let the agent run the command and inspect the durable result.

Use the Amics skill to recommend a GPU for this training command.
Do not acquire compute yet. I need at least 24 GB VRAM for about 30 minutes.

After reviewing the recommendation:

Approved up to $1.50/hour for a 30-minute command.
Run `uv run train.py`, keep the machine warm for at most 10 minutes,
and stop if the run becomes unknown.

What the skill teaches

• Recommendations are read-only; acquisition is a separate approval decision.
• Non-interactive runs need explicit price and runtime bounds.
• Source snapshots exclude common secrets and never replace the local tree.
• Environment values are forwarded only when named with --env.
• Auto-acquired compute stops by default; warm reuse is a bounded lease.
• Interrupted or ambiguous runs are inspected, never replayed automatically.
• Durable workspace creation and deletion are separate spend and data-loss decisions.

What agents should not do

Agents should not install themselves on the remote machine, forward their login or source-control credentials, turn a recommendation into implicit spend approval, delete a workspace without explicit intent, or replay an unknown run whose command may already have executed.