Commands and skills
Slash commands and $skills in mini-tui: what each command does, how skills are matched, how they reach the agent, and where they live on disk.
2 min read ยท Edit this page
Commands
Type / at the start of a message.
| Command | What it does |
|---|---|
/new |
Start a new chat. In the web app nothing is created until you send. |
/model [id] |
Open the model picker, or switch straight to an id. |
/compact |
Summarize the conversation now to free context. |
/subagents [name] |
List the subagents this session started, or open one (web app). In the terminal, /resume opens one. |
/connect |
Connect a provider with your own API key. |
/settings |
Open settings. |
/skills |
Browse the skills you can reference with $name. |
/help |
Commands and keys. |
Picking a command from the menu fills it in and never runs it, so a stray Enter cannot fire /new. A path
such as /etc/hosts, or a sentence that happens to start with a slash, goes to the agent unchanged.
Skills
A skill is a folder with a SKILL.md. Type $ anywhere in a message and mini-tui lists the skills, filtering
on any word of the name in any order: $body and $body-pr both find pr-body.
follow $good-code, then use $better-ui and $pr-body
The agent receives each referenced skill's instructions ahead of your prompt, then your prompt as you typed
it. The transcript shows only what you typed. An unknown $name stays plain text.
Where skills live
mini-tui keeps its own folder, ~/.config/mini-tui/skills/ (override with MINITUI_SKILLS_DIR). At install
and at every startup, each skill in ~/.claude/skills that the folder does not have yet is copied in.
Copies are never overwritten, and a skill you delete is not re-imported. bun run sync-skills runs it by
hand.
Bundled skills
mini-tui ships its own skills and installs them into that folder at install and at every startup. A newer version replaces a copy you have not edited; one you edited or deleted is left as you made it.
$subagents: how an agent splits work across subagents of its own session (Rust agent): read the room (agent resources), size the fan-out to the free memory against a 10 GiB reserve (up to 100 live children), spawn them, keep working while they run, steer them with messages, and collect and check their results.$e2e: how an agent writes and runsmini-agent-rs e2ebrowser tests.