Agents, models and modes

Agents

Astra runs two coding agents: Claude Code and Codex. Both ship inside the app, so there’s nothing to install — you connect your account once in Settings › Connections and the app runs the bundled binary on your behalf.

They’re independent. You can connect one and not the other, and different spaces (or different tabs in the same space) can use different agents. Whichever you pick, the agent runs locally against the space’s checkout with the same file, terminal and tooling access.

Models and effort

Inside an agent you choose a model, and for most models an effort level — how much thinking the model does before it answers. Astra shows the two together as one picker; the list comes from the agent itself, so new models appear as the providers ship them rather than when Astra updates.

Effort runs from minimal through to the highest reasoning tiers. Astra’s defaults are deliberate: quality-grade models start at high, which is genuinely useful without being slow, and the top tiers stay something you reach for on purpose. Speed-tier models default to low effort, since speed is the reason you’d pick one. Some models have no effort setting at all, and the picker simply doesn’t show one.

You can set a default model and default mode for new chats in Settings › General › Chats, and choose which models appear in the picker at all.

Modes

Mode is the dial that decides what the agent is allowed to do. It sits in the composer and you can change it mid-conversation.

Mode	What it does
Ask	Conversational. Reads and answers; makes no edits.
Plan	Proposes a plan of action. Still makes no edits.
Build	Executes — writes files, runs commands.
Design	Build’s equivalent inside a design tab: edits design frames.

The useful habit is Ask or Plan first on anything non-trivial, then Build once the approach looks right. Ask is for understanding a codebase; Plan is for agreeing on an approach before any file changes; Build is for doing it.

Mode is per chat tab, so one tab in a space can be exploring in Ask while another is building.

Usage meters

The composer can show two gauges, for both Claude and Codex:

• A context ring — how full the current conversation’s context window is. When it fills, start a fresh chat or clear the one you’re in rather than fighting it.
• A plan usage meter — how much of your provider plan’s allowance the current window has used. You can turn it off in Settings › General › Chats.

Both read from your provider account. Astra doesn’t bill for model usage — you’re using your own Claude or Codex plan.