All cards
VerifiedOfficialAgents

Agent Pulse

A live pulse of your agents: every model request of the agents you connect, as bars on a shared timeline (duration, tokens, tool calls), plus status timelines, turn durations and how long each agent has been waiting for you.

by NeuroSquad examplesv1.2.0verified Oct 5, 2026

Screenshots

What it can do

The same list the app shows before install. A card can do nothing else: it runs sealed inside its own card.

Asked at install

  • See the agents on this canvas

    Low risk

    Their names, which tool they run, and whether they are working, waiting for you or finished.

    agents.read

Asked only when needed

  • Read and change cards connected to it

    Medium risk

    Notes, checklists, task boards and stickies you connect to it with an arrow.

    cards.connected

  • See token usage and costs

    Low risk

    How many tokens the agents on this workspace used and what they cost.

    usage.read

The card asks for these while you use it, and you can say no.

Internet addresses

Connects to no internet addresses.

Reviewer’s note

Reads agent status (agents.read). Optional usage.read, asked when the Requests view opens: model request times, tokens and tool names of arrow-connected agents only, via agents.timeline (1.2: subagent requests drawn apart from the same data, app 0.1.270+). Never reads agent output. No network, no files.

— glmn-ai

What “Verified” means

The NeuroSquad team read this card’s source at the commit above, checked that the build matches it and that every permission is needed. The app installs exactly those files and refuses anything else. It is a review, not a guarantee — the card is still community code, so read what it asks for.

More cards

All cards
For authors

Built a card? Put it here.

Publish it in a public GitHub repository and open a pull request to the verified list. We read the code at one exact commit — then everyone can install it in a click.

Submit your cardCard SDK