Plan
One goal becomes a sprint.
HiveRunner turns an objective into a scoped sprint, splits it into owned tasks, and keeps acceptance checks and owners visible before agents start moving.
Goals → Sprints → Tasks
Turn a goal into an automated AI-agent sprint.
HiveRunner is the command center for running AI-agent teams: it breaks a goal into a sprint, splits the sprint into owned tasks, runs them through your agents, and keeps every run reviewable. Local-first by default — it runs on a machine you own.
Operator surface
live state, local control
HVR-323to-do
Assemble decision packet
HVR-319moving
Audit public claim surface
HVR-321review
Review hero composition
run streamlocal
09:42goalLaunch plan approved
09:43agentScout claimed HVR-319
09:44taskHVR-319 → in progress
09:45reviewHVR-321 awaiting review
goalOperator approves the plancontrolled
taskAgent work moves through statesvisible
reviewClosure waits for a humansafe
Product proof
A command center for people running real agent teams.
HiveRunner is built around one loop: define the goal, plan the sprint, split it into tasks, track the runs, and review before anything closes.
Run
Execution stays inspectable.
Follow task state, run output, artifacts, and review handoffs across the sprint instead of waiting on a black box.
Review
Nothing ships unreviewed.
Agent output moves through review states — never auto-accepted — and decisions, memory, and context carry into the next sprint.
operating loop
goal → sprint → tasks → runs → review
Goalowner review
Define the objective
Outcome, acceptance checks, and owner are visible before HiveRunner plans anything.
Sprinttasks + agents
Plan it, split the tasks
One goal becomes a scoped sprint of owned tasks, routed to the right agents.
Reviewhuman gate
Track runs, then sign off
Heartbeats, artifacts, and comments stay attached; nothing closes until you review it.
09:42:16goaloperator accepted the sprint plan
09:43:02taskHVR-319 moved into progress
09:44:31runartifact and review notes attached
First run
Start with a real team instead of an empty shell.
HiveRunner can launch a workspace with starter agents, public-safe role files, bundled portraits, and saved voice choices. Operators can review the team before anything starts running.
- Choose a starter team during workspace setup.
- Bundled avatars render immediately across the app.
- Saved voice choices are attached without blocking setup.
- Provider keys remain optional until the operator connects them.
Starter identities
The team arrives with faces, voices, and working style.
These are bundled agent identities, not placeholders. A new operator can start with a team that already feels distinct, then edit or regenerate anything later.
Vega
Brand / visual systems director
creative voice ready
Toby
UX / product analyst
product critique
Gator
QA / verification lead
review cadence
Prism
Frontend designer
interface craft
Overseer + Symphony
Let an AI operator supervise the work, not just answer prompts.
In HiveRunner, an overseer is a Codex or Claude agent that can watch the board with you: checking stale runs, following review cards, waking the right worker, and telling the human what needs a decision.
The idea is aligned with OpenAI's Symphony: coordinate agent work through shared state, plans, tasks, and reviews instead of treating every agent as a one-off chat window.
Watch
active runs and blocked cards
Coordinate
workers, reviewers, handoffs
Report
what needs human judgment
Honest signals
What this actually is.
HiveRunner is an open-source local-first command center, not a hosted SaaS. The code, license, CI status, and security reporting policy live on GitHub.
Product
See the operator model, runtime visibility, and control surface.
Go deeper on goals, sprints, tasks, runs, hives, and the local-first boundaries that keep agent work understandable.
Explore product →Docs / Quickstart
Clone the repo and run the local-first app.
Copy one setup block, review requirements, and understand what is required versus optional before starting.
Start locally →