Works with the agent tools and inference providers you already use
hob is where model choice, terminal workflow, and task context stay under your control while you run parallel agent work at full speed.
Give each task its own tiled workspace. Keep agents, terminals, specs, and diffs together so parallel work stays legible instead of turning into shell-tab sprawl.
Every agent run stays searchable and resumable. Stop re-briefing models from scratch just because you switched tasks yesterday.
Capture follow-ups while context is still warm. Keep real work attached to the workspace instead of scattering it across notes and throwaway tools.
Pro adds encrypted remote access from your phone, tablet, or browser without moving your workflow into a bundled provider cloud.
hob is built for professional developers who already use AI agents as part of real engineering work and want control over model selection, routing, history, and spend.
Bring Claude Code, Codex, OpenCode, and your own providers. Pay hob for the local-first workspace that makes serious agent work organized and portable.
Buy hob for employees or company-managed engineering workflows without forcing everyone through one bundled inference business.
Keep routing, access, and compliance inside your environment while giving teams a durable workspace for serious agent workflows.