Nodal supervises Claude Code and Codex across isolated worktrees and human-gated integration. Your agents work in parallel, and their work comes back as a branch you review before anything merges.
Request a beta inviteAgents never edit your working tree. Session agents work in Nodal-owned git worktrees, pipeline agents in a staged clone outside the repo, and their work reaches you as a branch and a pull request.
Transcripts, approvals, run records, staged transfers, and committed work are durable across reconnects, worker restarts, crashes, and project reopens. A crash ends the live agent process, and Nodal tells you exactly what stopped rather than pretending otherwise.
Merging is a human action that no autonomy setting can grant. Context sharing, spend and destructive calls pass through explicit gates, and approval is bound to the exact action and material you reviewed.
Nodal drives your logged-in Claude Code and Codex sessions, so runs bill the subscription you already pay for. It does not resell inference.
Start a session, delegate a goal, or open a Roundtable. Each agent gets controlled territory in its own worktree.
Inspect progress, costs, requests, conflicts, and failures. Pause, retry, stop, or resume without losing the run.
Inspect the output, run verification, and approve the commit from the Map's commit inspector. An agent can open a pull request, but merging is yours alone.
Put Claude Code and Codex in the same room. Nodal runs structured positions, critiques, revisions, and comparison, then shows accepted points, unresolved disagreements, and evidence linked to the exact frozen context each model received.
Multiple models review the work. You make the call.
Refresh the token on every request where expiry < 60s. Simple, and it never serves a stale session.
That fires a refresh per in-flight request. Under concurrency you get a token stampede and the last write wins. Refresh once behind a lock.
Resume managed work or start from a goal.
Delegate an autonomous run with visible limits and controls.
Structured cross-model critique with frozen context.
See agents, branches, dependencies, locks, and contention.
Rule on every gated action from one audit surface.
Track spend per project and per run against a monthly budget.
Pipeline agents run under a macOS seatbelt write fence confined to their staged run folder. Session agents run in Nodal-owned git worktrees rather than under the OS fence, because confining them would mean admitting the shared .git directory and everything in it. Reads stay broad and the model CLIs reach the network. Nodal keeps its state on your machine and records activity in a hash-chained event log whose head is stamped into the git commits it writes, so the history can be checked by something other than Nodal.
Session agents are not OS-fenced. Reads are broad and the model CLIs reach the network, so a run can read anything your user account can. The monthly budget tracks spend and warns, it does not block it; the enforced ceilings are the operator's. Old audit entries age out under a retention window. And if you turn on API-key mode for the operator, your Anthropic key is stored on your machine, encrypted by the macOS keychain, and goes nowhere except Anthropic.
Claude Code and Codex produce work. Nodal coordinates, isolates, supervises, reviews, recovers, and integrates that work.
Beta 1 is free, invite-only, and built for macOS on Apple silicon. It currently supports Claude Code and Codex through their existing CLI sessions. The first cohort is limited to 20 to 50 developers.
Request a beta invite