Choose theme
AI Resources
OpenHands Agent Canvas
OpenHands Agent Canvas is a self-hosted workspace for running coding agents and automations across local machines, Docker containers, servers and cloud backends.
It supports OpenHands, Claude Code, Codex, Gemini and other ACP-compatible agents. You choose both the agent and the backend that gives it access to your project. Use this as a first read, not a recommendation. Open the original project before trusting details like terms, limits, privacy, cost, setup, or safety.
What it is
Keep coding work visible
A dependency-update agent on a team server and a personal coding session on your laptop can appear in the same Canvas interface. They can use different backends, credentials and project folders.
Why it stands out
Choose the agent and its execution environment
Canvas is the control interface; an Agent Server runs the work. Switching from OpenHands to another ACP agent is a different choice from moving execution into Docker or onto a VM.
Availability
Public source; beta project
The current flagship OpenHands repository contains Agent Canvas and documents local, Docker and self-hosted paths. Its README labels the project beta; model access and connected services have their own requirements.
Why it matters
What makes it useful
If code review should continue while your laptop is closed, Canvas can connect to an Agent Server on a separate server. That changes where the work runs; the interface on your laptop is no longer the machine that must stay awake.
What to know
Where it fits
Canvas sits above the Agent Server and automation service. The server supplies agents, tools and conversations; the automation service handles schedules and webhook triggers. Running the frontend alone does not supply a coding runtime.
Notable points
What stands out
The current README documents a Docker container for each new conversation. Existing local conversations are not converted. Containers preserve workspace files and conversation state, and two conversations using the same host workspace still share its files.
Before using
What to review
The direct local installation gives the agent server access to the host filesystem. Docker access depends on the mounted projects and configuration.
Per-conversation Docker execution does not put the whole Canvas or automation service inside that sandbox.
Remote hosting needs the project's security setup, including authentication, HTTPS and network restrictions; a VM is not an access policy by itself.
The self-hosting guide notes that the bundled editor shares Canvas's browser origin. Code running on that origin can read stored session keys for all registered backends; a separate backend does not isolate those browser-held keys.
Check the selected agent's credentials and model requirements, and review edits before accepting or publishing them.
Reader fit
Who may find it relevant
Developers who need one visible workspace for several coding agents or execution backends.
Teams connecting review or maintenance jobs to a server and scheduled or event-driven automations.
Someone wanting a simple coding chat may not need to host an Agent Server and automation service.
Editorial note
Why LifeHubber lists it
Canvas can run several coding conversations with separate Agent Servers and tools. For parallel edits, choose separate project directories or Git worktrees when the files should stay apart: separate containers can still mount the same host workspace.
Source links
Source materials
Reader note
Before relying on this entry
LifeHubber lists entries to help readers inspect AI projects, not to endorse them or prove they are safe, suitable, accurate, maintained, or right for a specific use. We do not verify every entry in depth. Before relying on anything listed, review the original materials, terms, privacy practices, limits, and risks that matter for your situation.
What to explore next
Keep the coding workflow understandable and recoverable.
Once one workspace can switch agents and backends, the next questions are what should stay in your own files, how other agent setups organise the work, and which permissions matter when an agent can use the desktop.
More in AI Agents
Keep browsing this category
Explore more AI agent projects.
Paperclip
paperclipai/paperclip
A self-hosted server and dashboard for coordinating agent teams through companies, goals, roles, issues, heartbeats, budgets, approvals, and persistent activity records.
OpenSEO
every-app/open-seo
An SEO workspace with keyword, ranking, backlink, and site research tools, an MCP connection for AI agents, reusable research skills, and hosted or self-hosted access.
Emil Kowalski Skills
emilkowalski/skills
Design-engineering skills for AI coding agents, with separate paths for building motion, reviewing animations, finding worthwhile opportunities, and naming the effect you want.
For project maintainers
Listed here? You can use the badge.
If you maintain a project with a current LifeHubber listing, you may add the optional “Listed on LifeHubber AI Resources” badge to its README, docs, or website. No introduction or permission request is needed.
