Workflows

Routines that run themselves,on a canvas you can see.

A workflow in NeuroSquad is just cards and arrows: something to start it — an agent’s daily prompt, a web watch that sees a change, a message to your Telegram bot — agents to do the work, terminals and browsers as their tools, and notes for the output.

A morning brief, a release started from Telegram, and a dependency update started by a web watch.

At 07:30 two agents get their own daily prompt: one reads the analytics dashboard, the other reads the repo. Both write into one note; the Standup card shows what each did.

Visible wiring

The canvas is the workflow. No YAML to keep in sync.

Three ways to start

A daily prompt on the agent card, a web watch that sees a change, or a message to your bot.

Queues that wait their turn

Prompts queue per agent and drain when the current turn is done.

Guard rails

Permission prompts, isolated worktrees and a budget brake.

Workflows

Anatomy of a routine

Trigger → work → check → output. Every step is a card.

01
Trigger
Daily prompt, web watch or Telegram.
02
Work
Agents with terminals and browsers.
03
Check
A reviewer agent or your approval.
04
Output
A note, a Telegram message, a tag.

Common routines

Morning briefReleaseDependency updatesChangelog watchTest triageContent drafts
Beyond code

Agents for the rest of the business

The same cards run inbox, sales, back-office and automation work. Each page below is a live, clickable scenario.

Draw your first routine

Three cards and two arrows are enough to start.

The first-run wizard installs an agent CLI, Git and the dictation model for you — or skips each step.