What is Gonu?
An AI workspace for developers and engineering teams. Chat and source-based drafts run on the web; authorized local file work uses Worker or another supported client.
Start with one useful taskGONU / DOCUMENTATION
Small steps. Clear examples. Work you can review.
Learn what each tool does—and how to put it to work.
Find the right tool.
See how it works.
Try it with your context.
Start with a repository and one focused task.
Turn source notes into a reviewed team handoff.
THE TOOLBOX
What it does. How to use it. What to check.
34 guides for real work
Connect repository structure, engineering decisions, incidents and selected documents. Inspect evidence before planning, changing or recovering work.
Inspect real checks, protect acceptance rules, correct memory, cap task usage and review alternatives before choosing a result.
Go from a small source document to a draft you can check and reuse.
Think through an idea, explain pasted code, or work with a specialist in a shared conversation.
Give a reusable specialist a clear job, instructions and a limited set of tools.
Save a repeatable procedure that an agent can apply when it fits the task.
Share an agent or skill deliberately, or install a reviewed public item into your own library.
Organize documents and tasks under the right project and audience.
Import selected text and ask for answers tied to those sources.
Turn an authorized transcript into a reviewed summary, decisions and trackable actions.
Start from a useful request structure for repository explanations, onboarding, meetings, updates and summaries.
Find the request, selected evidence, output and actual status of a web workspace task.
Bring an authorized local repository into a task with file access, diffs and approved checks.
Choose the published Linux package that matches your distribution and CPU.
Explore project structure, relevant files and important execution paths before making changes.
Choose how much work a task should do and which actions require your decision.
Inspect proposed file changes and use recorded snapshots for deliberate recovery.
Inspect shared engineering sessions, recorded phases, approvals and validation evidence from the web.
Review an authorized issue request and follow an isolated implementation through approval gates.
Connect an allowed GitLab project and review issue work through the shared execution flow.
Inspect actual browser assertions, screenshots and logs from an isolated application check.
Connect a supported tool server, discover its tools and enable a deliberate selection.
Choose among Gonu-provided models available to your plan and workspace.
Use a supported model already installed on your machine through a local client.
Give existing Gonu users the right workspace and project access.
Keep team conventions and decisions available in the right scope, alongside recorded changes.
Check your account’s current allowance, subscription status and available upgrade path.
Use shared tasks and approved local repository tools from a terminal.
Work with selected code, project context, native diffs and shared sessions inside the editor.
Use one Gonu account across supported surfaces and keep workspace access separate from device sign-in.
Find your referral link and inspect the referral activity recorded for your account.
Make reviewed Gonu-hosted model offers available to eligible users.
Use the platform’s restricted operational pages to investigate accounts, usage and customer requests.
Find the existing site-audit, metadata and content tools in the platform administration area.
WALK THROUGH IT
Synthetic sample content, with simplified screens. No model call, upload, account change or paid task runs here.
Open Knowledge. Paste a short text or import a plain-text/Markdown file. Give it a descriptive title. Mark practice content as an example.
On Home, select Summarize documents, choose your imported source and an available model. Read and accept the selected-source sharing consent.
Open Tasks & history. Read the status and source markers, check important statements against the source, then export Markdown if useful.
A REQUEST TO START WITH
Summarize this project brief. Separate the goal, constraints and unanswered questions. Cite each source and do not invent missing details.
THE ESSENTIALS
Setup, access and a few useful boundaries.
An AI workspace for developers and engineering teams. Chat and source-based drafts run on the web; authorized local file work uses Worker or another supported client.
Start with one useful taskChoose an available installer for your OS and architecture. Review release notes and publisher information; keep operating-system protections enabled.
Read the Worker guideUse an available AppImage or distribution-compatible package. Match the published architecture and follow the package manager’s dependency guidance.
Follow the Linux guideChoose an enabled model for your plan. You do not need a personal API key for Gonu models. Optional local inference and existing workspace connections have different setup and data flows.
Choose a modelOpen an authorized folder in Worker. Ask “Explain this project”, check the file references, then plan one small change. Review writes and commands before approving.
Explore repository understandingImport selected text, choose a template and model, approve processing of those sources, then inspect the result and its references in Tasks & history.
Learn selected-source tasksChoose a shared workspace, grant existing users appropriate roles and organize project access. Use shared rules and memory for recurring team context.
Set up team accessYour live account shows the current offer and allowance. Model multipliers affect usage. A payment redirect alone does not activate paid access.
Understand usage and billingCLI and VS Code are source/local beta surfaces. They reuse Gonu identity and shared task APIs. Check access and distribution requirements before installing.
Explore developer setupBEFORE YOU START
No. Every walkthrough is an interactive illustration with synthetic content. It does not call a model, upload a file, create an assignment or use your allowance. Use the clearly separate product link when you are ready for real work.
Not for Gonu-provided models. Your platform administrator configures those providers; you choose among the models enabled for your plan. Local Ollama and existing workspace-key connections have their own setup.
No. GitHub, GitLab, MCP and browser execution require appropriate configuration and permissions. Their guides explain prerequisites. An example screen is not a claim that your server has that integration enabled.
No. A local model runs inference on your machine, but Gonu orchestration and session history can still receive selected task context. Read the local-model guide before sharing private material.
These guides cover the active developer and business workspace. Retired creative and recruitment workflows are not promoted here. Historical content and existing account entitlements remain separate.
Bring one source, ask one useful question, and check the answer.