Terminals ​

The terminal at the bottom is a real shell opened in your workspace. Focus it with ⌘2 / Ctrl+2 and open a new one with ⌘T / Ctrl+T. Tabs include:

  • Agent — where the agent runs its commands, visibly, so you always see exactly what it did.
  • Your own terminals — add as many as you like with +; use them like any terminal.

The agent can open its own dedicated terminals for interactive processes (REPLs, ssh, msfconsole) and drive them — sending input and reading output — and it can read your terminals too.

Command history & @terminal ​

Each terminal records the exact commands you typed, not just raw scrollback. That powers two things:

  • @terminal in chat attaches the terminal you're focused on — its recent commands plus scrollback. Use @terminal:<id> to attach a specific one.
  • The agent can recall the exact commands you ran on demand, so questions like "what did I just type?" resolve cleanly instead of guessing from output.

You can also select any terminal output → "Add to chat" to hand the agent a precise snippet.

Safety

Destructive commands (e.g. rm -rf, sudo, git push --force) pause for your approval before they run — see Approvals & safety.

Downloads the agent triggers are saved into your workspace (under hs-output/) so you can open and inspect them.

See also ​

Hacker Sidekick documentation