Console Agent Session Control: Product Model and Design Rationale

Status: Design rationale behind Console Agent Session Control (Phase 2, Agent Orchestration Program) — the tree-view, instance-row actions, and hardline integration described here have shipped; this page keeps the product-model reasoning that motivated the shape, out of the roadmap item's day-to-day tracking.

Problem

jackin❯ operator escape hatches used to leak Docker details. When the operator wanted a shell in a running environment, the practical path was a raw command such as:

docker exec -ti jackin-chainargos__agent-brown zsh

When the operator wanted to reconnect to the original foreground agent, the path was jackin hardline <container-name>. That worked for one primary process, but it gave the console no first-class model of "this workspace or directory has one isolated container with multiple active agent sessions inside it."

The missing product model was a running container as an isolated collaboration environment, with one or more agent sessions inside it. The operator needed to be able to open jackin console, see all running instances for configured workspaces and ad-hoc directories, see each agent role and exact agent session inside those containers, attach to one, start another, open a shell, or hardline the original agent without remembering Docker names.

Goals that shaped the shipped shape

• Operators can open a shell in a running role container from CLI and console without typing docker exec.
• Operators can see which configured workspaces and ad-hoc directories have running or preserved instances.
• Operators can see which agent sessions are running inside each instance.
• Operators can reconnect to the primary session through the console.
• Operators can start and reconnect to a secondary agent session inside the same container.
• Exiting one agent session does not stop the container or other concurrent sessions.
• Raw Docker container names are no longer required for normal shell, hardline, or session attach flows.

These goals are why the console models a container as a tree node with expandable instance-session children, rather than a flat list of containers or a single "attach" action — the container is the collaboration environment, and the tree exposes the sessions living inside it as first-class rows with their own action set (reconnect, new session, shell, stop, purge). See the current shipped state and any remaining work in Console Agent Session Control.