Taskboard

Tech

Alias

Screenshots

Model

AI provider

Approval mode

Roles

⚒ coder every project — works todo cards

🔬 reviewer policy only — review cards get picked up; nothing starts when you toggle this, and turning it off does not stop a reviewer session already running

🧭 tech lead policy only — briefs a card before a coder builds it. Inert until a tech-lead token exists on an enabled env (a machine self-mints its own once its env/fleet policy allows it); turning it off does not kill a lead session already running. ⚠ tech-lead is fleet-wide: enabling it can activate the grooming gate for every project that uses tech-lead, not just this one 📚 knowledge curator policy only — curates project docs and memory. Inert until a machine's env/fleet policy enables knowledge-curator (it self-mints its own token then); turning it off does not kill a curator session already running ⚡ executor not decorative like the other three — this is the gate a fired `prose` action run checks: off here 422s the fire itself instead of just queuing quietly. Still policy-only otherwise — it mints and starts nothing on its own

🌐 served by

orchestrator claude-orchestrator

bridge offline — on the hub run: systemctl --user restart taskboard-orch-bridge (unit not installed? see "Orchestrator chat" in the README)

Tech

Alias

Screenshots

Model

AI provider

Approval mode

Roles

AI Routing

Docs

💰 Spends

🏠 Home

📥 Backlog

📁 Projects

Environments

Why is this blocked?

⚙ Settings

👤 People

🤖 Workers

▦ Fleet Slots

Fleet slot

New task

New project