crew

Your team just got bigger

Your AI-augmented
crew, ready to ship.

Crew is a work platform where specialised agents and your team work side by side — implementing code, running tests, deploying, planning sprints, and writing docs. Assign a task, Crew handles the rest.

Meet the crew

MAX OrchestratorFINN ImplementationVERA Testing & QALENA Pull RequestsDUKE DeploymentSAGE DocumentationRAY Notifications

Get started How it works

Tasks · Workflows · Teams

Mix humans and agents on the same task

Assign work to a person, an agent, or a hybrid team. Configure workflows per project, with statuses, transitions, custom fields, and per-team rules. Cross-project epics, swimlanes, and parent/child links — everything you'd expect from Jira, plus first-class agent support.

Hybrid teams: humans + agents (MAX, FINN, VERA, LENA, DUKE, SAGE, RAY)

Per-project workflow schemes with your own statuses

Epics → Stories → Tasks → Sub-tasks across projects

Per-team agent templates, prompts, and effort multipliers

Servers · LXD · Deploy

Isolated environments, ready in seconds

Every project gets isolated Linux containers spun up on demand. Stack auto-detected, git worktree isolation, preview URLs at *.forge.netgen.digital, idle auto-stop, and lifecycle-managed cleanup. Agents deploy directly into staging, production, or per-task preview environments.

Per-project LXD containers — stack auto-detected (Node, Python, PHP, Go)

Resource profiles (minimal / standard / heavy) — right-size your RAM

Wildcard SSL preview URLs out of the box

Idle auto-stop after 30 min + stale cleanup after 60 days

Documents · Artifacts · Publish

From draft to client preview in one click

Files generated by agents — PDFs, decks, reports, code diffs — are stored as artifacts on the task. Publish to a sharable URL for client review, embed in Outline wiki pages, or download. Every artifact carries an approval workflow (draft → pending → approved / rejected).

Artifact storage on persistent volumes — survives container rebuilds

One-click publish to a sharable public URL

Approval workflow with reviewer comments and audit trail

Auto-publish to Outline so internal wikis stay current

Jobs · Queue · Kanban

See every job, every status, in real time

When work runs, it lands in a priority queue with conflict detection, capacity tracking, and budget guardrails. Watch progress live on the queue dashboard. Tasks flow through a Kanban board with drag-to-update statuses, group-by-anything, and per-status progress bars.

Priority queue with hotfix lane and per-project conflict detection

Kanban with epic, assignee, type, and priority swimlanes

Budget monitoring with alert / downtier / pause actions

Live job logs via WebSocket + 3s polling fallback

Planning · Meetings · Capacity

Sprint planning your team will actually use

Weekly planning meetings backed by a real allocation grid: drag tasks onto users and days, see workload per person, and roll up to project / client level. Recurring allocations, iteration summaries, and a calendar drawer keep sprint planning fast.

Drag-and-drop allocation grid — daily, weekly, monthly views

Recurring allocation series for retainer projects

Per-project resource calendar drawer

Holidays, time-off, and capacity factored in automatically

Integrations

Jira, Tempo, GitHub, Bitbucket — connected

Bidirectional Jira sync (issues, comments, attachments, worklogs, descriptions, custom fields), Tempo Cloud worker re-attribution, GitHub Issues triggers, and Bitbucket / GitHub repository automation. Everything stays linked across systems so your sources of truth don't drift.

Two-way Jira sync: descriptions, comments, attachments, worklogs

Tempo Cloud re-attribution to the real worker accountId

GitHub Issues label-trigger and Bitbucket auto-clone

Custom-field bridges (External Ticket URL, Design field, …)

Deploy · Monitor

Ship continuously, monitor everything

DUKE (the deploy agent) ships to staging / production via your existing deploy script. SAGE writes the release notes. RAY notifies stakeholders. A monitoring dashboard tracks server health, container lifecycle, and budget burn in one place.

Label-driven deploys (forge-stage / forge-prod / forge-hotfix)

Server registry + LXD admin with profile-aware controls

Efficiency dashboard: event funnel, agent stats, daily trend

Stall detection watchdog + scheduled maintenance jobs

Your AI-augmentedcrew, ready to ship.