Claude Code and Codex, one shared workspace

Your team and its AI coding agents. One workspace.

Fluxius runs Claude Code and Codex in shared, persistent sessions — with a team-lead agent, approvals, runners, integrations and usage control built in.

Web app · Telegram bot · desktop and mobile apps in development

fluxius · Payments API · Fix webhook retries

A real recording of the Fluxius web app with demo data. Interface labels are translated into English for this page.

Works with the tools you already use

Claude CodeCodexAnthropicOpenAIOpenRouterZ.aiGitLabGitHubBitbucketJiraLinearYouTrackTrelloSentryGrafanaMetabaseConfluenceNotionFigmaTelegram
The product

Everything happens in the session — together.

A session is a shared channel between people and agents. Pick a stop and watch the real interface.

Real Fluxius, demo data

Teammates see what the agent does as it happens. Presence, “need help”, a message queue and sessions that persist — so nobody works alone, and nothing is lost when a tab closes.

The agent asks before it pushes
Human in the loop

Agents are fast. You stay in control.

Run in approve-dangerous mode and risky actions pause until someone with write access says yes — from the web, Telegram or your phone.

  • Approve, deny, or allow for the session. Everyone in the session sees the request; every decision is recorded.
  • Budgets that stop runs. Monthly token and cost limits per project, person or agent — warnings at 80%, a hard stop at 100%.
  • An audit log of logins, agent and secret changes, approvals and tool calls.
  • Read freely, write after approval. Integrations let agents read on their own; comments, branches and MRs need a human yes. Merging is reserved for the team-lead agent, and only with a green pipeline and no conflicts.
The platform

Everything around the agent, already built.

Runners, tools, integrations, accounts, usage and access — so teams can ship with agents without building the plumbing.

Runners that fit the job

Several runners per project, each with its own image and size. Run them on Kubernetes or Cloudflare Sandbox, or connect your own machine. Hours are metered against your plan.

KubernetesCloudflare SandboxYour own runnerImage catalogSizesOutbound proxy
Project settings, Runners tab: a runner with its image, labels and connection instructions

A workspace in the browser

VS Code, a terminal, live preview of your ports, generated artifacts and — when the runner has it — a secure desktop (VNC), all in a panel next to the chat.

Start from the source

A task, an error or a merge request from a connected service becomes a session with its full context. Agents read your tools directly — and write only after a human says yes.

JiraLinearYouTrackTrelloGitLabGitHubBitbucketSentryGrafanaMetabaseConfluenceNotion
Integrations catalog with Jira, Linear, YouTrack, Trello, GitLab, GitHub and more

Bring your own accounts

Connect Anthropic, ChatGPT (device login), OpenRouter or Z.ai accounts per project. A run takes the first account that is under its limit and moves on when one runs out. Keys are encrypted and never shown again.

Project settings, Models tab: provider accounts for Claude, Codex and OpenAI, OpenRouter and Z.ai

Usage you can read

Tokens and cost by project, person, model and day — plus how much of each subscription window is left, and monthly budgets that stop runs at the limit.

Usage tab with token and cost totals, a daily chart and a breakdown by person and model

Numbers that live on Home

Agents answer with cards like this one. Pin them to your home screen as widgets, then drag, resize and publish them to the project.

An infographic card in the chat: delivery stats and a daily chart

Pick the agent per session

Give each project agents with their own avatar, prompt, models and skills — then choose one when you start a session, and keep a session private if it’s just for you.

New session dialog: choose between the Backend agent and the Reviewer, pick a model, optionally make the session personal

A sidebar that scales

Search, filter to “only mine” or by status, sort, go compact, pin favorites, reorder by drag. The master session always sits first, marked with a crown.

Sidebar view settings: scope, filters, sorting and display options

Agents that wait and recheck

Scheduled tasks on a cron, Advisor mode for autonomous goals, and wake-ups so an agent can come back to a result later.

Warm prompt cache optional

Keep an active session’s prompt cache warm so the next turn is faster and cheaper. Off by default.

Roles and private sessions

Owners, developers and viewers per project, with Keycloak groups. Keep a session personal, then share it read or write.

Figma, both ways

Agents read your designs and comments without a plugin. With the Fluxius plugin open they can change the design too — and every change can be undone from the project’s Figma tab.

Telegram and Mini Apps

Talk to agents and answer their questions from Telegram. Building a Telegram Mini App? Test it inside the Preview panel with an emulated Telegram client and fake init data.

Your language, your theme

The interface comes in Russian and English, in light, dark or system theme with your accent color. Agents answer in the language you choose, in every session.

The Fluxius session view in the light theme

Your agents, your tools

Use Fluxius from your own Claude Code, Codex or Cursor through an API token and an MCP endpoint — with your permissions, narrowed by scope.

Account settings: API tokens and a snippet to connect Claude Code, Codex or Cursor

Memory that follows you

Agents remember how you like to work — across projects and sessions. It’s yours alone: other members, admins and other people’s agents never see it, and you can export or erase it any time.

Notifications that find you

Browser notifications and Telegram when a run finishes or fails, when an agent needs approval or asks a question, and when a spending limit is close.

How it works

From task to merge request in four steps.

1

Set up a project

Connect repositories, add secrets and MCP servers, pick a runner and plug in Jira, Sentry and GitLab.

repos payments-api
runner backend-full
secrets ••••••• encrypted
2

Configure an agent

Pick models, write the system prompt, add skills and sub-agents, set permissions.

model claude · codex
skills review, release
perms approve-dangerous
3

Start a session

Blank, or straight from a ticket, an error or an MR. Invite teammates, or let the team lead start it.

source jira PAY-312
branch fluxius/stop-retrying-410
people Anna, Max
4

Review and ship

Check the diff, approve risky steps, push and open a merge request.

+4 −1 retry.ts
24 tests passed
MR !482 opened
Claude Code (plain terminal) vs Fluxius

Same agent power. Team-sized workflow.

Fluxius doesn’t replace your coding agent — it runs Claude Code and Codex for you inside a shared workspace and adds what a terminal on one laptop can’t.

Plain terminal (Claude Code, Codex CLI)
Fluxius
Who’s in the session
One person, one terminal
Everyone on the project, live — presence, “need help”, a shared queue
History
Lives on one machine
Persistent shared sessions anyone with access can continue or fork
Harnesses
One agent per terminal; switching means starting over
Claude and Codex in the same session — switch mid-conversation
Coordinating work
You juggle terminals and tabs
A team-lead agent starts sessions, collects reports and merges MRs
Dangerous actions
A prompt in your own terminal
Any teammate with write access approves; every decision is audited
Where the task comes from
You paste context in
Start from a Jira or Linear task, a Sentry error or a GitLab or GitHub MR
Environment
Your laptop
Managed runners with images, sizes, VS Code, preview, VNC and artifacts
Spend
Whatever your own account shows
Usage by project, person, model and day — plus monthly budgets that stop runs at the limit
Away from your desk
Tied to the machine you started on
Open the web app anywhere, get notified when an agent needs you, or answer in Telegram

Each session works in its own git worktree, so parallel tasks never step on each other.

Security and control

What an engineering lead asks before saying yes.

Approvals

Risky actions wait for a person with write access. Codex runs in a sandbox, and Fluxius asks before MCP tool calls.

Encrypted secrets

Project secrets and provider credentials are stored encrypted and stay out of the frontend and the logs.

Audit log

Who logged in, who changed an agent or a secret, who approved what, and every tool call the agents made.

Budgets

Monthly token and cost limits per project, person or agent; owners are warned at 80% and runs stop at 100%.

Roles and private sessions

Access follows your Keycloak groups. Personal sessions are invisible to everyone, admins included, until the author shares them.

Isolated worktrees

Every session gets its own branch and working copy. Agents get no more access than the token you give them.

Wherever you work

Web today. The rest on the way.

The web app is the product. Desktop (Tauri) and mobile (Capacitor) apps are thin shells around the same interface and are still in development.

Assistant outside the project in the works

A personal assistant that isn’t tied to a single project’s repository — for the questions that don’t belong to one codebase.

IDE plugins in the works

Bring Fluxius sessions into your editor, next to the code.

FAQ

Questions, answered.

Does Fluxius replace Claude Code?

No. Fluxius runs Claude Code (via the Claude Agent SDK) and Codex (via the Codex SDK) for you in a shared workspace. You keep the agents; Fluxius adds sessions, collaboration, approvals, runners, integrations and usage control around them.

Which agents and providers are supported?

Two harnesses today: Claude Code and Codex, behind one runtime interface. Provider accounts can be Anthropic, OpenAI (including ChatGPT login), OpenRouter or Z.ai. Other terminal agents are not supported yet.

Can I use Claude and Codex in the same session?

Yes. A session keeps its own history, so you can switch the model — including between Claude and Codex — and the next turn continues with full context. Both work on the same worktree, and every turn shows which model produced it.

What does the team-lead agent do?

The master session is a regular session whose tools are for managing other sessions: it lists them, starts new ones with a task, sends messages, interrupts and archives. Work sessions report back to it, and it can merge an MR under the project’s rules. It never gets more rights than your project settings allow.

What can an agent do without asking?

That is configured per agent. In full-access mode it works autonomously. In approve-dangerous mode, Claude Code pauses risky actions until a teammate with write access approves. Codex has no per-command prompts in its SDK, so it runs inside a workspace-write sandbox and Fluxius asks for approval on MCP tool calls.

Where do agents run?

On runners: managed ones on Kubernetes or Cloudflare Sandbox, or your own machine. A project can have several, each with its own image and size; usage hours are metered against your plan.

Which tools does it connect to?

Jira, Linear, YouTrack and Trello for tasks; GitLab, GitHub and Bitbucket for code and reviews; Sentry for errors; Grafana and Metabase for data; Confluence and Notion for docs; Figma for designs — plus any MCP server you add and a Telegram bot. A task, an error or an MR from a connected service can start a session with its full context. More services are on the catalog’s “soon” list.

Where can I use it?

In the browser today, plus a Telegram bot. Desktop (Windows, macOS) and mobile (iOS, Android) apps are in development as thin shells around the same web interface.

How do I sign in and who sees what?

Sign-in goes through your company’s Keycloak. Access is by project, using groups for admins, owners, developers and viewers. Personal sessions are visible only to their author until shared.

Ready when you are

Give your team an agent everyone can steer.

Create a project, connect your repos and start your first shared session.