The control center can bring its own model and can host several agent implementations through ACP-compatible backends.
Agent Atlas / Coding agents
OpenHands/OpenHands
A self-hosted developer control center that can run OpenHands and other ACP-compatible coding agents across local, Docker, remote, and cloud backends.
Coding agent platform
Full dossier
Project documented
Reviewed August 26, 2026
Intended operator
Individuals and teams hosting persistent coding agents and automations.
Primary surfaces
Agent Canvas / agent and automation servers / local, Docker, remote, and cloud backends
Tracked repository
Teaching mechanism
The executor boundary changes with the selected backend
The same control surface can connect to an agent running directly on a host, in Docker, on a VM, or in a cloud service; the README explicitly warns that an unsandboxed server has full filesystem access.
Evidence limitDeployment options and warnings do not establish equivalent isolation or operational guarantees across backends.
Inspect the source
System Map
Eight inspection boundariesConversations, project mounts, automations, and backend-owned state shape durable work.
Agent Servers execute work on a selected local, container, VM, remote, or cloud environment.
Isolation depends materially on backend choice; direct host execution and Docker mounts carry different authority.
Always-on servers and switchable backends make continuity a service-level concern.
Automations and issue decomposition coordinate work; agent-specific worker semantics depend on the selected backend.
Agent Canvas centralizes conversations, backend choice, automations, and team operation.
Architecture docs, deployment commands, warnings, backend APIs, tests, and future source reporting expose the control plane.
Learn Through This Project
10 lessons
Chapter 1
Before the First Model Request
Chapter 2
The Invisible Prompt
Chapter 3
Tool Calls: Contract, Authorization, Execution
Chapter 4
Loops, Workflows, and Recovery
Chapter 5
Delegation Is a Distributed System
Chapter 6
Memory Is a Typed Storage Contract
Chapter 7
The Approval Trap
Chapter 8
The Human Control Loop
Chapter 9
Trust Must Make Its Case
Chapter 10
The Future Has Three Branches
Latest Reporting
6 articles
A Completed Agent Run Is Not a Successful One
September 1, 2026 — OpenHands now treats a finished process and a successful task as different facts—so an automation can be completed, blocked, and operationally failing at the same time. OpenHands/OpenHands
Coding Agents Are Building Themselves—But Git Can Prove Only Part of ItAugust 11, 2026 — Across 40,345 unique commits in ten public agent codebases, coding-agent credit becomes sharply more visible in 2026—and the missing receipts matter as much as the names that appear. openai/codex, google-gemini/gemini-cli, anomalyco/opencode, QwenLM/qwen-code, cline/cline, aaif-goose/goose, OpenHands/OpenHands, charmbracelet/crush, mistralai/mistral-vibe, Kilo-Org/kilocode
The Supervision Derby finish: which coding-agent repo made control more real?July 31, 2026 — The finish is a split decision: Codex makes filesystem policy travel into capability discovery; Qwen Code makes a fork’s tool gate survive resume; OpenHands and Gemini CLI make narrower but useful controls reachable. openai/codex, google-gemini/gemini-cli, QwenLM/qwen-code, OpenHands/OpenHands
OpenHands: separating the software agent from the sandbox it works inJuly 30, 2026 — A new healthy-local-backend fallback makes the project’s larger argument concrete: the thing that helps a person steer an agent must stay legible even when the agent is running somewhere else. OpenHands/OpenHands
The Harness Was Taking the Test TooJuly 30, 2026 — OpenAI says two context choices lifted GPT-5.6 Sol from 13.3% to 38.3% on the ARC-AGI-3 public set. Codex source shows those “settings” are really a production memory system—and rival harnesses reveal how much of it can be copied. openai/codex, pydantic/pydantic-ai, earendil-works/pi, NousResearch/hermes-agent, mistralai/mistral-vibe, anomalyco/opencode, QwenLM/qwen-code, google-gemini/gemini-cli, cline/cline, aaif-goose/goose, OpenHands/OpenHands
The Supervision Derby: four coding-agent control surfaces enter the first Diff DerbyJuly 29, 2026 — The opening call gives Codex a narrow provisional edge for making subagent instruction precedence survive delegation and resume—but Qwen Code, Gemini CLI, and OpenHands each put a different kind of operator control on the track. openai/codex, google-gemini/gemini-cli, QwenLM/qwen-code, OpenHands/OpenHands
Primary Sources
Open Questions
- Can the UI make the effective backend authority and isolation level impossible to miss before work begins?
Evidence state describes what this Atlas profile can support today. It is not a product score and it is not evidence that uninspected capabilities are absent.