Cloud Agent runs coding agents against a user's repository inside retained sandbox containers. Users work from a web interface, while enrolled hosts run Claude Code, Codex, pi, OpenCode, or Kilo. The control plane coordinates jobs and sessions, and relays the agents' own pushes and pull requests to GitHub. Accounts are its own: people sign up, apply through a waiting list and are approved by an administrator. Every task runs on a model connection its owner adds with their own provider key, such as a FreeInference key.
This is the documentation for the software itself: running it locally, operating a deployment, understanding how it works, and contributing to it. The source is at github.com/HarvardMadSys/hybridInference-cloud-agent, under the MIT license; bugs and questions go to its issue tracker.
| If you want to | Start here |
|---|---|
| Run the browser, API, hosts, and sandboxes on one machine | Local Docker stack |
| Set up a checkout, find the code that owns a change, and run checks | Developer guide |
| Find the module that owns a responsibility and follow its dependencies | Code structure |
| Change the owner web app, and see the rules and APIs behind each feature | The owner web app |
| Send a change | Contributing |
| Translate a page, or update a translation after an English edit | Translating the docs |
| Configure credentials, enroll hosts, qualify Kata, deploy, and migrate | Operations runbook |
| Upgrade an existing deployment and see what each migration changes | Upgrade notes |
| Look up a configuration name, default, or owning component | Configuration reference |
| Build release artifacts and deploy them in the required order | Releases |
| Understand the sandbox trust model before deploying | Security |
| See the API this service serves, as its committed contract | API contract |
These pages document the software, not any one installation of it. How end users work with a particular deployment, which agents and models it offers, and how to get an account there are the operator's to publish. Dated designs, plans, and acceptance records stay in the repository as engineering history and are linked from here where they explain a decision.
:caption: Getting started
:maxdepth: 2
local-stack
development
code-structure
contributing
translations
:caption: Concepts
:maxdepth: 2
context-recall
agent-access-policy
local-accounts
api-contracts
security
web-app
:caption: Operations
:maxdepth: 2
operations
upgrade-notes
configuration
operations-compute
releases
deployment-reliability
disaster-recovery
runtime-settings
host-execution-limits
runner-lease-troubleshooting
smoke