Type / at the start of an empty composer and a picker opens. It holds two different kinds of thing:

• Skills — packaged instructions that shape how the model does one job.
• Commands — /goal, /loop, /followup and /compact, which change how the chat behaves rather than what the model is told.

Navigate with ↑ ↓, choose with Enter or Tab, dismiss with Escape. Add your own text after the skill name before sending.

A skill is a SKILL.md file: a short front matter block and a body of instructions. Sending /skill-name loads those instructions and pins the skill on the chat for subsequent turns.

Agents can also choose relevant skills during requested work with Load skill. The tool discovers enabled skills, loads their instructions, and reads bundled text references. It does not pin a skill or execute its steps; the agent uses its existing tools and permissions to follow the workflow. Skills marked disable-model-invocation: true remain available through explicit slash invocation and are excluded from agent discovery.

Built-in skills are enabled by default.

Turn individual ones off in Settings → Integrations → Skills. Only enabled skills appear in the picker.

Some skills are model-invocable and some are not. A skill marked otherwise can only be run by you typing it, so the model cannot decide on its own to enter caveman mode.

Settings → Integrations → Skills → Skills Library browses curated third-party packs: Matt Pocock, Addy Osmani Agent Skills, Superpowers, last30days and Browserbase. Install a whole pack or pick individual skills. You can also install directly from a GitHub URL.

Each pack is pinned to a specific commit rather than tracking a branch, so an install is reproducible and cannot change under you. Downloads are restricted to GitHub hosts. Installs land in skills/ in your Minnow home, record where they came from, and are enabled immediately.

Browsing works offline — Minnow ships an index of each pack's contents. Installing needs a network.

Create a folder in skills/ under your Minnow home with a SKILL.md inside:

---
name: release-notes
description: >-
  Draft release notes from the git log since the last tag. Use for /release-notes.
disable-model-invocation: true
---

# Release notes

1. Run `git log <last-tag>..HEAD --oneline`.
2. Group commits into Added / Changed / Fixed.
3. Write one user-facing line per entry. No commit hashes, no internal ticket ids.
4. Call out anything that breaks compatibility under its own heading.

Then enable it in the Skills catalog. A user skill wins over a built-in of the same name, so you can override a shipped skill by shadowing it.

Set disable-model-invocation: false (or omit it) to let agents discover and load your skill during relevant work. Keep it true for workflows you want to start explicitly.

/goal <condition> tells the chat to keep working until something is actually true — not until the model feels finished.

After each turn a separate evaluator agent checks the condition against the code and test results, then either confirms it or sends the chat back to work. It is the difference between "I have fixed the tests" and tests that pass.

• /goal all tests in test/orchestrator pass
• /goal clear stops the loop (stop, off and reset also work)

The goal persists on the chat across reloads. Bind a capable model to the evaluator role in Models → Routing → Goal evaluator — a weak evaluator will happily rubber-stamp a broken build.

/loop re-runs a prompt on a schedule while Minnow is open and the chat is idle.

The chat panel shows a countdown with pause, resume, skip, interval edit and stop. Loops expire after seven days.

A global ticker wakes at each loop's stored due time, so a reload or a laptop sleep does not lose the schedule.

/goal, /loop and /followup are mutually exclusive on one chat, and /clear clears all three.

/loop is not the Scheduler. A loop lives in one chat, keeps its context, and needs that chat idle. A Scheduler job is a headless run in a chosen workspace with a chosen model, independent of any conversation. Iterating on something belongs in a loop; a nightly report belongs in the Scheduler.

/followup hands the work on to a new chat instead of continuing this one. The new chat opens with a context summary of the chat before it plus the task it exists to do, so nothing has to be re-explained.

/followup 5 review the build for bugs and fix them starts with the bug review, then hands off four more times — each link choosing the next task from what the link before it did. Chains stop at ten links.

The first follow-up opens in front of you because it is your own next task. Later links arrive in the sidebar with an unread dot, so a long chain does not keep pulling the window away. The chat panel shows how many links are left and has a Stop chain button; /followup stop does the same. Arming a chain never sends anything to the model, and it is safe to type while the current reply is still running.

A chain is stored on its chat, so it survives a reload or a restart: quit with a link pending and it fires when Minnow comes back.

The agent-chosen tasks use the model bound in Models → Routing → Utility tasks.

/compact folds older turns of the chat into a summary now, instead of waiting for the window to fill. The last couple of turns stay word for word, and nothing is deleted from the transcript. Text after the command tells the summary what to keep: /compact keep the schema decisions. /compress and /summarize are aliases. See Context, memory, and rules.

• Working in chat
• Modes
• Scheduler app
• Settings app