Introducing bbasata Cloud: coordinate your AI coding tools

Today we are thrilled to announce the launch of bbasata Cloud — a shared coordination layer for teams that build software with AI coding tools.

The Problem: AI Adoption Is Fragmented

Real teams do not all standardize on one AI tool. One developer prefers opencode, another Claude Code, another Kilo. The problem is not a shortage of agents — it is the lack of a shared layer to coordinate work across all of them.

• Prompts and rules live scattered across each tool's configuration.
• Tasks move from one AI to another manually, losing context on the way.
• There is no shared view of who did what, with which permissions, and with what result.
• The more the team relies on one provider, the more the whole workflow depends on that single vendor's roadmap.

The bbasata Cloud Solution

bbasata Cloud does not replace your AI tools. It coordinates them. Projects, plans, phases, sprints, boards and tasks live in one place, alongside handoffs, agents, permissions and a full audit trail.

The core model is handoff-first: bbasata packages each task into a portable work packet — context, objective, relevant files and current state — that any team member can continue from their favorite AI tool. Work never gets trapped in a single chat session, editor or vendor.

Here is what makes bbasata Cloud different:

• Bring your own AI coding tool — Works with opencode, Kilo, Claude Code, Antigravity, Devin and Goose, plus any MCP-compatible client. Exporters generate native agent configuration per client.
• Handoff-first coordination — Work is packaged into portable packets, so a task can move from one tool to another without losing context.
• Governance for teams — Roles, granular allow/ask/deny permissions per agent (bash asks by default; edit, git, webfetch and websearch configurable), and an audit trail of every action.
• Agile built in — Plans, phases, sprints and boards are first-class, not a bolt-on. Visual, durable workflows connect the work of agents with the real project flow.
• No prompt leaks — Agent prompts are served via MCP Resources at runtime; bbasata agents sync writes only native client configuration (model and tool permissions) to disk, never prompt bodies.

What Agents Are Available?

At launch, bbasata Cloud offers agents across six industry categories:

• Marketing & Growth — SEO specialist, copywriter, social media manager, email marketing specialist, growth hacker
• IT & Development — Product manager, UI/UX designer, frontend developer, backend developer, devops engineer, QA engineer, security engineer, data engineer, technical writer
• Content & Creative — Novelist, screenwriter, editor, translator, creative writer
• Legal & Compliance — Legal advisor, compliance officer, contract analyst
• Education — Tutor, curriculum designer, evaluator
• Technical Support — Level-1 support, level-2 support, technical account manager

How It Works

Getting started takes about two minutes:

1. Create your account — No credit card required. Accounts are free; every organization starts with a 7-day trial.
2. bbasata init — One command creates the team config file.
3. bbasata login — Paste your API key to link your team.
4. bbasata agents sync — Exporters generate native setup for opencode, Kilo, Claude Code, Antigravity, Devin and Goose. Tasks appear on the shared board as portable handoff packets.

Pricing

We believe in transparent pricing. No per-token surprises. Each plan includes monthly calls and agents; every plan starts with a 7-day trial, no card required. See pricing for current limits. Visit our pricing page for current details.

Built for the MCP Ecosystem

bbasata Cloud is built on the Model Context Protocol (MCP), the open standard for AI-client integrations. This means it works with any MCP-compatible client out of the box. We are not building a walled garden — we are building a service that plugs into the tools you already use.

Because prompts are served via MCP Resources, your API key grants access to agent capabilities without ever exposing the underlying instructions. This is a more secure approach than storing prompt files locally.

What is Next?

This is just the beginning. Over the coming months, we will be adding more agents, more categories, and more advanced features including multi-agent workflows, custom agent requests for Business subscribers, and deeper analytics.

We are also open-sourcing our MCP client and validation packages. The entire ecosystem is designed to be transparent and extensible.