Gleipnir Get access

Control room · Claude Code · v0.1.0

Run more
agents than
you can watch

Gleipnir puts every Claude Code session on one panel. Which are working, which are blocked waiting on you, and which one has been rewriting the same file for twenty minutes.

Request beta access

macOS · universal · needs Claude Code installed

Fleet · 8 agents Program · ios-client

The problem

Two agents fit in two terminal tabs. Six do not.

Past a handful of sessions you stop knowing which ones are moving. You alt-tab to find the one asking a yes-or-no question, and while you look for it another one has quietly spent four dollars retrying the same failing test.

Gleipnir is one window holding all of them, grouped into projects you can close and reopen with the whole crew intact. It runs the real claude CLI in a pseudo-terminal — the same sessions, the same transcripts, resumable from a plain terminal if you ever stop using it.

What is on the panel

Eight modules,
one window

Each one is built on something Claude Code already writes to disk. The source is printed on the right, the way a rack unit is marked with its input.

Fleet

Every session as a tile: working, idle, or blocked on a human. It also finds sessions Gleipnir never started — including the one in your terminal.

~/.claude/sessions/*.jsonpolled at 1 Hz

Strain

Watches the last sixty tool calls of a running agent for thrashing — the same file, the same test, over and over — and warns before it becomes an invoice.

transcript tail, 512 KBnever a full parse

File wall

One column per agent, files stacked beneath, reconstructed from what each agent actually did. In a shared working copy git can tell you a file changed; it can never tell you which agent changed it.

~/.claude/projects/<slug>/*.jsonlworks with no repo at all

Usage

Real token counts, per agent and per project, priced against the current table. Counted from the transcripts, not estimated — though on a subscription the dollar figure is a size signal rather than a bill.

transcript usage recordsinput · output · cache

Advisor

Reads those same numbers back to you in a sentence. A four-hundred-thousand token turn is worth knowing about; you should not have to go looking for it.

arithmetic onlyno extra model call

Jobs

Ask a question without paying for the conversation. A live session re-sends its whole history every turn; a one-shot starts empty, so “did the tests pass” costs a couple of thousand tokens instead of hundreds of thousands.

claude -pruns whether the agent is up or not

Crew

Send one instruction to several agents at once. Anything not running is reported back by name rather than silently dropped.

broadcastcollisions flagged first

Remote

A local server and an iPhone client. Watch the fleet from the couch, answer the agent that is blocked, and put your laptop down.

HTTP + WS on your LANtoken paired, off by default

On screen

What it looks like
with six agents up

A demo fleet, not a real one - two projects, six agents, work in flight.

Six agent tiles. One is blocked on a question, one is flagged as straining after editing the same file nine times, the rest are working or parked.
Every session at once. API is blocked on a yes-or-no question. UI has been rewriting one file for nine attempts and is flagged for it. DOCS is parked and costing nothing.
One agent large in the centre with a live terminal, the rest alongside as smaller tiles.
The one you are working with is front and centre and fully interactive. The rest keep running beside it.
A column per agent, files stacked beneath each, with one file appearing in two columns and flagged in both.
The same fleet by what it is changing. billing.ts is in two columns at once, flagged in both - git can tell you a file changed, not who changed it.
A small floating panel of square tiles, one per agent, showing state only.
Shrink the whole window to a panel of states and put it in a corner. Click any tile to open that agent and talk to it.

How it reads Claude Code

No API.
Just the files.

Claude Code writes a heartbeat while it runs and a transcript as it works. Gleipnir reads those. That is the whole mechanism, and it is why the app can see a session it did not launch.

None of it is a supported interface. It can change in a Claude Code update, so every parser here is built to degrade to empty rather than throw: an update should cost you a feature, not the app.

Gleipnir reads your local files and spawns the CLI you already have. There is no account and no telemetry. Nothing of yours leaves the machine — the only thing it asks the network is whether a newer build exists.

The Gleipnir mark: a bound wolf.

The name

Named for a ribbon.

Two iron chains couldn't hold Fenrir. A ribbon could — not by making the wolf weaker, but by giving the gods something to hold him with.

That's the job here. Eight agents aren't too powerful, they're too many to watch. Gleipnir puts all of them on one panel and tells you which one has gone off the rails.

Download

Version 0.1.0

Gleipnir is in closed beta while the rough edges get found. Ask and you will almost certainly get in — it is gated to keep the feedback readable, not to be exclusive.

macOS · Apple silicon and Intel Closed beta

Request access

Say roughly how many agents you run at once. You get a download link and a code by reply.

Size
175 MB
Requires
macOS 11 or later · universal
Signing
Developer ID, notarized

Windows and Linux Coming soon

Both are in progress. Each installer has to be produced on its own machine — the terminal module is native and cannot be cross-built — so they arrive separately rather than all at once.

Before you open it

Gleipnir drives the Claude Code CLI you already have. Install it and sign in first, or the app will open to an empty rack.

claude auth status