AI Coding Agent Dashboard for Real Execution Work
Mrrlin is an AI coding agent dashboard for goals, tasks, inbox decisions, run status, artifacts, and review evidence.
Chat tools vs Mrrlin
The execution layer beats another blank chat window.
Chat tools make you re-send context and manage the work. Mrrlin keeps project memory and spends tokens deliberately.
Your agents, under one harness
One coding agent gives you output. Mrrlin gives you a harness.
Codex, Claude Code, and other agents can write. Mrrlin keeps the brief, routes each task, challenges weak work, and brings you in only for decisions that need you.
Read the workflow transcript
Without a harness, a person carries context between prompts, checks each output, follows up, and decides what happens next. With Mrrlin, one goal becomes a visible plan: work is routed, reviewed, retried when weak, and brought back only for the decisions that need a human.
Why it matters
Watching agents is easy. Knowing which one needs you is the hard part.
Mrrlin gives operators a dashboard for what agents are doing, what is blocked, what changed, and what needs approval.
Workflow
From opening the board to clearing everything that needs you.
Scan the board
Every task carries a status, from backlog and to do through in progress, awaiting verification and handoff needed to done, and the Kanban board shows where each one stands. Plans show how far each goal has come, and the Autopilot view lists what is up next, what is blocked and which task is holding it up.
Open any run
A run displays where it stands, from running through self-review, pull request and dev or prod deploy, along with the model provider behind it, its branch and everything it recorded. Test results and the diff summary are read where they were captured, not in a separate log.
Clear the inbox
Questions, approvals and handoffs from every task collect in one inbox. Answer in your own words, or let the Director decide, and the tasks waiting on that answer continue without another prompt from you, in the order their dependencies allow.
Act where you look
Approve a launch, send work back for another pass, or ask the Director to unblock stalled handoffs and continue, all from the view you used to find them. Watching and deciding happen in the same place instead of two.
Use cases
Where it fits.
Agent run monitoring
Mid-afternoon, the view shows three runs in progress, one in self-review and one failed on a type error. The failed run's error artifact sits on the run itself, and its task is back in the queue with that output attached for the next attempt.
Blocked question triage
Four tasks are stalled on questions, among them whether the new filter should include archived projects and which date format the API returns. Answer them in one sitting from the inbox, and each task resumes as its answer arrives instead of waiting for you to visit its terminal.
Build evidence review
Before trusting a change to the checkout flow, open its run and read what it recorded: the suite's results, the commands it ran, the diff summary and the preview address it produced. The evidence is the run's own record, not a summary written after the fact.
PR readiness
A pull request is ready when its run shows a green build, a reviewer's pass from a second model and no open questions. The AI coding agent dashboard puts those three side by side, so approving the merge or the preview deploy is a decision rather than an investigation.
Comparison
A status screen, or a dashboard you can act from.
Keep reading
More on coordinating coding agents.
Start from the AI coding agent orchestration overview, or go deeper with the guides next to this one.
FAQ
Before you start.
Is Mrrlin just an AI coding agent dashboard on top of my agents?
No. The dashboard is one surface of it. Underneath, Mrrlin plans the work, runs each task in an isolated worktree, has a separate model review the output, requires a receipt to close a task and writes decisions back to memory. It is the execution and quality-control layer around your agents; the view is how you watch it work.
Can I see which model did which piece of work?
Yes. Every execution run records the provider that performed it, along with its branch, base commit and artifacts, and reviewer notes and consensus results stay attached to the task. When a task went through several attempts or several agents, the history shows who did what and what was sent back, which keeps the work auditable later.
Does the dashboard track deploys as well as code changes?
Yes. Runs move through dev deploy and prod deploy statuses, every deploy attempt is kept as an artifact, and a produced URL shows where a preview landed. A task's deploy policy, off, preview or production, sets how far it may go, and the production step sits in your inbox as an approval before your repository's workflow is triggered.
Orchestrate Claude Code & Codex CLI without terminal lockups.
Stop juggling disconnected CLI windows and resolving manual merge collisions. Mrrlin coordinates your agents in isolated git worktrees, collects passing test artifacts, and opens verified pull requests.
No credit card · No migration · One goal