Arbiter is the OS-level orchestration brain that sits above your LLM clients and beside your tools. It dispatches and supervises your agents, terminals, and long-running jobs. It applies configured routing and scheduling policy, then preserves a clear record of work interrupted by a crash or power cut.
Agnostic by design. Bring configured clients and supported providers, or none at all. Public binary release preparation is underway.
$ arbiter run "refactor the auth module" --client claude-code --worktree
run f3a1 queued
run f3a1 routed → configured tier (policy selected)
run f3a1 deferred → off-peak window (acct-2 quota in 34m) ...
run f3a1 running worker 48211 · managed worktree
-- power cut · machine reboots -- $ arbiter hydrate
run f3a1 suspended state preserved, operator action required
[ok] 1 interrupted run accounted for, 0 lost
Illustrative. Arbiter is a working alpha; recovery behavior depends on job safety.
The problem
LLM clients run one session well, but they don't orchestrate many across processes, machines, accounts, and time. Memory tools remember and code tools verify, but neither dispatches. The orchestration layer is simply missing, so today it's run by hand.
Many clients, each good at one session: Claude Code, Codex, Gemini, Cursor, Windsurf. Running ten of them means ten terminals and a mental map you have to hold yourself.
Every provider has its own quotas and peak-hour throttling. Keeping work on whichever account still has headroom is a babysitting job nobody should have.
Long jobs sprawl across terminals, branches, and machines. The moment the power blinks or you reboot, the in-flight work is gone and the day with it.
What Arbiter does
Arbiter is the muscle and the dispatcher. It never runs the coding turns itself. It dispatches the clients that do, and owns everything around them: routing, accounts, scheduling, processes, and durable run accounting.
One command center for terminals, processes, git worktrees, work-groups, and sub-jobs, across Windows and modern Linux. Everything you launch, in one place to see and steer.
Apply configured capability, cost, and safety rules to routing decisions. The learned optimization layer is still being hardened for release.
Schedule configured work around availability windows and account constraints. Exact provider behavior depends on the clients and accounts you connect.
Use configured local and remote control surfaces, including supported messaging adapters, an HTTP control plane, editors, and MCP clients. Availability depends on your setup.
A crash or reboot leaves an accounted record of interrupted work. Safe read and inference jobs may be retried; side-effecting work stays suspended for review.
Configured credential pools and scheduling policy help select available routes. Treat account rotation behavior as setup-dependent during alpha.
Watch it run
No slideshow, no mockups. This is a running alpha daemon on a real machine: configured routing, live runs, approvals, and the local brain assisting the operation.
Silent screen capture, 8:02. Prefer YouTube? Watch it there.
The headline
The thing that ends most long AI sessions is not a bug. It's an outage: a reboot, a dropped connection, a blink in the power, and a day of in-flight work evaporates because nothing was holding it.
Arbiter holds the durable record. After a restart,
arbiter hydrate accounts for interrupted jobs, preserves
their state, and marks them suspended for deliberate operator recovery. Safe read and inference
jobs may be retried; side-effecting work stays held until an operator chooses what happens next.
$ arbiter status
3 runs were interrupted by an unclean shutdown.
$ arbiter hydrate
run f3a1 suspended state preserved
run 9c0e retried safe read task restarted
run b27d suspended needs approval to retry
[ok] 1 safe retry · 2 held for review
Illustrative output.
Model routing
Different jobs call for different capability and cost tradeoffs. Arbiter gives you a place to configure those choices, inspect routes, and keep safety policy in the decision path.
Set capability, cost, and safety policy for the jobs you dispatch. Arbiter records the route it selects.
The routing layer is designed to weigh cost alongside capability. Its learned optimization remains under hardening during alpha.
Routing decisions are recorded, and configured safety rules can constrain which routes are allowed for sensitive work.
Agnostic by design
Arbiter is built to run with zero providers and zero sibling tools, and to get sharper as you add them. Swap any client or provider out and it still runs. If it only had value when you ran the whole stack, it would have failed its own test.
Arbiter
The orchestration brain
Dispatches, routes, schedules, and records the state of work it supervises. This app.
Your LLM clients
Dispatch targets
Claude Code, Codex, Gemini, Cursor, Windsurf, or your own. Bring any, or none.
Your providers
Accounts & quota
Configured accounts can be selected with routing and scheduling policy in mind.
Your tools
Optional delegates
Memory and code-analysis tools plug in when present, and are never required.
Standalone-first. Integrations are additive, never load-bearing. Arbiter is the dispatcher above them all, not a member of the set.
Where this is going
One instruction, and Arbiter handles the dispatching, the accounts, the timing, and the recovery underneath it. These are the north-star behaviors the whole system is built toward.
“Run this across off-peak hours on whichever account still has quota.”
“Spin up 10 workers on this monorepo, each on its own git worktree; recover them if I reboot.”
“Text me on Signal when the suite goes green, and let me stop any run from my phone.”
“Index the repo, warm it with the project's standards, then dispatch the fix to whichever client I prefer today.”
The ecosystem
Five focused tools for shipping with AI agents. Each stands on its own; together they cover code, action, orchestration, memory, and proof.
TheAuditor
Don't let agents guess about your code.
theauditortool.com →Warden
Keep AI coding work visible across supported providers.
wardenclient.com →Arbiter
The command center for your AI operation.
you are hereCurator
Don't let agents guess about you.
curatormcp.com →BenchProctor
A public SAST benchmark built to resist label leakage.
benchproctor.com →● Working alpha · Release preparation
Arbiter is now in IP hardening, licensing, and public binary packaging preparation. The source remains private. Follow the engineering journal and the public release channel for verified progress and availability.
No signup funnel. Product status stays public through the journal and release channel.