What To Configure For The Agent
What To Configure For The Agent
Section titled “What To Configure For The Agent”Do not prepare every capability on day one. Pick the tier that matches what you want the agent to own now.
| Tier | Goal | You provide / confirm | Agent owns |
|---|---|---|---|
| 1. Local development | Code and debug locally | Repository entrypoint, local browser-debug boundary, optional auth_hint / debug_credentials | Service modeling, language runtime / deployment config, lifecycle, logs, browser interaction, breakpoints, authentication |
| 2. Remote debugging | Investigate remote bugs | Host, remote superdev-agent install method, remote service and log hints | Remote deployment / log source config, cross-host logs, diagnostics, evidence collection, root-cause reasoning |
| 3. Team deployment | Deploy and roll back | Pipeline config and variables; human-managed ingress, SSL, DNS | Pipeline validation, deploy, rollback, run logs, artifacts |
How To Choose
Section titled “How To Choose”- Only local coding and debugging: prepare tier 1.
- Need production-like logs or remote machines: configure tier 1 plus tier 2.
- Want releases and rollback through the agent: add tier 3.
The tiers are additive, not competing choices.