timmy-tui Run demo

Local-first agent trust console

Trust the receipt, not the model.

TIMMY TUI helps developers run AI agents with visible model choice, inspectable MCP evidence, browser/workspace review, and sealed local receipts.

Install TIMMY Watch the tour
$ npx timmy-tui demo
$ npm install -g timmy-tui
  • 0.1.0npm published
  • npxclean demo verified
  • localreceipt first
TIMMY TUI hero screen showing the trust receipt launch message and terminal UI.
TIMMY mascot peeking over a terminal window.

Friendly, fast, in the terminal

TIMMY walks the user through every proof state.

The mascot gives the trust workflow a clear emotional anchor without hiding the serious product: every command still resolves into evidence you can inspect.

Cards lift on hover
Receipt rails brighten
Proof pulse follows run state

Mascot browser commercial

A 15-second spot for the feature people remember.

This ad gives TIMMY a real character moment: mascot first, browser companion second, and the product feeling lands before the viewer has to parse a technical workflow.

mascot hook browser feature launch-ready ad

Proof rails

Every run gets a cockpit.

TIMMY is not another black-box agent surface. It is a review layer for the work agents do: model choice, tool permissions, workspace changes, and receipt fields stay visible.

01

Choose the brain before execution.

OpenRouter model selection stays visible before a run starts, so routing is an explicit choice.

02

Turn tools into files you can inspect.

MCP servers become command plans, permission outlines, receipt fields, and local evidence bundles.

03

Open the workspace without losing the terminal.

Browser companion and workspace views mirror the run while the terminal remains the source of truth.

04

A local record for what changed.

Receipts summarize run ID, prompt, tools, changes, artifact paths, and manifest hash.

Show the agent, not just the answer

The UI explains what the run can prove.

The launch page uses real product screens: model rail, MCP bundle, workspace companion, and receipt proof. It is designed to feel useful before it feels decorative.

TIMMY workspace and browser companion screen.
Workspace + Browser Companion
TIMMY OpenRouter model selection screen.
OpenRouter model rail
TIMMY MCP to CLI evidence bundle screen.
MCP to CLI evidence
TIMMY sealed receipt screen showing verified receipt fields.

Sealed local receipts

Make the proof feel alive.

TIMMY’s receipt model turns the run into a reviewable artifact: what happened, which tools ran, which files changed, and where the evidence lives.

Run IDrun_demo_178094...
Toolsfs.read, fs.write, cmd.exec
Statusverified

Interactive Arcade demo

Watch the launch flow before you run it.

The guided tour covers model choice, MCP evidence, workspace review, sealed receipts, and the final install command.

OpenRouter rail -> MCP evidence -> workspace review -> sealed receipt -> install CTA
demo.arcade.software / New Product Launch Highlights

New Product Launch Highlights

OpenRouter, MCP evidence, workspace review, receipts, and install CTA.
Open in Arcade
TIMMY Cloudflare roadmap screen showing Workers, R2, D1, Durable Objects, Vectorize, and Pages.

Local first. Cloudflare next.

Start local. Add hosted proof when the workflow earns it.

TIMMY’s first promise is local proof. Cloudflare becomes the natural next layer for receipt intake, artifact archives, indexes, workspace state, and semantic recall.

Install now

Clone TIMMY, build it, then start the TUI.

Use the GitHub repo for the full source path. Use the public npm command when you just want the fastest Product Hunt demo.

$29 launch deal

Builder Starter Pack

Setup notes, field guides, examples, and the upgrade path for teams building with agent receipts.

$250 deposit

Custom MCP proof pack

Send an MCP server URL and get a TIMMY-style command plan, permission outline, and receipt field bundle.