Open sourceTypeScript · ≈94k stars · MIT
OpenHands
The open platform for cloud coding agents.
OpenHands (formerly All Hands AI) is an open-source coding agent platform — agent canvas, sandboxed execution, SDK and an enterprise control plane — with 500+ contributors and millions of downloads.
Pros & cons
Each point cites its sources and says what kind of evidence backs it. A second reviewer tried to refute every claim before it was kept.
Pros
- User reportsSandboxed by default (Docker, with Kubernetes or VM options), so agents get file and bash tools without touching the host.[1][2]
- MetadataSelf-hostable and MIT-licensed, which suits teams that need an open alternative to hosted agents.[1][2]
- User reportsUsers praise its Software Agent SDK for driving agents from code in semi-deterministic workflows.[1][2]
- User reportsUsers value the public OpenHands Index benchmark for comparing models.[1]
- Best for
- Teams that want an open, sandboxed coding-agent platform they can self-host and drive programmatically.
- Watch out for
- Budget for container infrastructure, because it is heavier to run than single-binary CLIs.
Assessed Oct 29, 2026 · adversarially reviewedSuggest a correction
Activity
- Last push
- · today
- Latest release
- v1.43.0
· yesterday - Last 90 days
- 20+ releases
- · yesterday
1.43.0 (2026-11-20) What's Changed Features feat(automations): create automations enabled from the setup wizard by @SaltGardenia in https://github.com/OpenHands/OpenHands/pull/19769 feat(settings): Features page for experimental user feature flags, with Conversation reactions…
- · 3d ago
1.42.0 (2026-11-18) What's Changed Features feat(cloud): pick the exact connected git provider for repositories by @tofarr in https://github.com/OpenHands/OpenHands/pull/19762 feat(verify-openhands): every run has a non-loopback plain-HTTP origin by @enyst in…
- · 5d ago
1.41.0 (2026-11-16) What's Changed Features feat(verify-openhands): scenarios beyond the happy path, map controls, browser timezone by @enyst in https://github.com/OpenHands/OpenHands/pull/19670 feat(acp): choose the reasoning effort from what the agent reports by…
- · 9d ago
1.40.0 (2026-11-12) What's Changed Features feat: improve Codex subscription authentication guidance by @xunjizhe in https://github.com/OpenHands/OpenHands/pull/17194 Bug Fixes fix(chat): keep messages at tied history page boundaries by @enyst in…
- · 10d ago
1.39.0 (2026-11-11) What's Changed Features feat(automations): filter the dashboard by owner with a Service accounts option by @hieptl in https://github.com/OpenHands/OpenHands/pull/19472 feat(chat): move the message hover timestamp closer to its message by @ak684 in…
Checked against GitHub on Nov 21, 2026. Release notes are shortened to plain text.
The team behind it
San Francisco, CA · Est. 2024
The open platform for cloud coding agents.
OpenHands (formerly All Hands AI, born from the OpenDevin project) builds the major open-source coding-agent platform — agent canvas, sandboxed execution, SDK and an enterprise control plane — now rebranded under the project's name.
In research (18)
- Harness Leverage Across Model Size: A Ladder of Five Local Agent Harnesses from 0.75B to 4.7BVerdict: Useful
Nov 16, 2026 · Its StuckDetector is cited as the precedent for the duplicate-call intercept.
Nov 16, 2026 · One of the two agentic harnesses evaluated, compared directly against the authors' SimpleAgent.
- FeatureArc: Benchmarking Coding Agents on Evolving FeaturesVerdict: Important
Oct 27, 2026 · Open coding-agent platform, cited in the paper. FeatureArc-style multi-stage evaluation applies directly to it.
Oct 26, 2026 · One of the harnesses in PAUSE, with a different tracked-file trigger rule.
- Work Agents: Systems, Capabilities, and OptimizationVerdict: Incremental
Oct 26, 2026 · Agent execution harness in the survey's harness table.
- CoSA: The Conflict of Skills Attack in Real-World LLM Agent EcosystemsVerdict: Important
Oct 26, 2026 · Evaluated harness, with the highest violation rate under attack (66.11%).
Showing the 6 newest of 18 papers.
Neighbours on the workbench
Shell · ≈329k stars · MIT
mattpocock/skills
Small, composable agent skills for real engineering work.
Maintained by Matt Pocock (AI Hero)
MDX · ≈323k stars · MIT
obra/superpowers
An agentic skills framework and software methodology that works.
Maintained by Jesse Vincent (obra)
JavaScript · ≈308k stars · MIT
affaan-m/ECC
Agent harness performance optimization system: skills, instincts, memory and security for coding agents.
Maintained by Affaan Mustafa
Connections
Relations from the Harness Hub knowledge graph, as charted in the State of Harness reports.
Maintained by · 1
- OpenHandsStartup
Similar entries · 2
- OpenHandsStartup
- SandcastleProject
