Your AI Development Team, Everywhere You Work

Thu, 05 Nov 2026 00:00:00 GMT

Starting today, your BoldCreate AI development team works in the cloud, on a schedule, on your machine, and in your terminal. Cloud Agent keeps approved tasks running after you log off, Automations start tasks for you, the Desktop App runs locally, and the CLI works in your CI pipelines.

Enterprise Server gives admins one place to manage people, budgets, and models.

Every surface ends the same way: a change with tests and review findings that your team reviews and merges. Your organization’s skills, approval policy, and budgets apply wherever the work runs.

Important: Syncfusion Code Studio is now BoldCreate. We’ve retired the Code Studio IDE. If you already have a Code Studio license, you can move to BoldCreate and keep using the credits remaining on your account. Read the Code Studio to BoldCreate migration guide to see what changes and how to move your team.

Cloud Agent

Cloud Agent runs your AI development team in the BoldCreate cloud, on your existing repositories. Each task works in its own isolated sandbox, so parallel tasks don’t collide.

Start a task

Choose GitHub, GitLab, or Gitea, pick a repository or an environment, and describe the change. Pick the agent for the task: Agent builds and tests a change, Ask answers questions without editing code, and Plan writes an implementation plan first.

Pick the repository and the agent, then describe the change in plain language.

Review the finished task

The agent keeps a to-do list as it works and reports what it changed and how it checked the result. The Changes panel shows every changed file, and Create pull request sends the branch to your repository for review.

Read the summary, open the diff, and create the pull request when you're ready.

Reuse an environment

An environment describes the workspace once: its repositories, a setup script for tools you install once, variables and secrets, and who can use it. Every task that starts from it gets its own copy.

Each environment shows its status, scope, and repositories.

Add plugins, skills, and MCP servers

Customize lists the plugins installed for you, your team, or your organization, plus a marketplace of more. Plugins add skills, agents, rules, and MCP servers in the cloud and on your machine. Integrations connect your Git provider and add an organization bot that comments on pull requests and starts automations.

Install a plugin once, and its skills and tools are available to every task.

Automations

Automations start Cloud Agent tasks for you, so recurring work such as fixing failing checks, reviewing pull requests, or updating dependencies doesn’t wait for someone to remember it. Each run works in its own sandbox and appears as a task in your sidebar.

Start from a recipe

Recipes cover common jobs, such as Fix failing CI, Review a pull request, Find vulnerabilities in a pull request, Weekly dependency update, and Investigate a new ticket. A recipe fills in the trigger and the instructions, and a new automation stays off until you turn it on.

Choose where the recipe works, then save it.

Build your own automation

Run an automation every hour, day, weekday, or week, or with a custom cron expression. You can also start it from pull request and check events, or from a webhook URL that your CI, monitoring tool, or ticket tracker calls.

Use Test run to try an automation before you turn it on.

Desktop App

The Desktop App puts your AI development team on your machine, in one window, for Windows and macOS. Choose This computer, and the agent works in your real folder, with your local database, uncommitted changes, toolchain, and the app running on localhost. Nothing is uploaded unless you say so.

Run a task on your computer

Pick a folder, describe the change, and send it. The agent explores the project, explains what it found, and lists each step as it works.

Local tasks and cloud tasks share one task list.

Approve each command

In Manual mode, the agent asks before it runs a command or changes a file. Open Tool Input to see the exact arguments, then allow or deny it.

Each request shows the tool and its arguments before anything runs.

Review every change

The Changes panel lists every changed file with its line counts. Open a file to read the diff next to the conversation before you create a pull request.

Open any file to read its diff next to the conversation.

Switch modes or send work to the cloud

Switch a task you trust from Manual to Auto, and the agent runs commands and changes files without asking. Choose Cloud instead of This computer, and the task runs in a cloud sandbox that keeps going after you log off.

Manual asks before each command or file change. Auto runs without asking.

CLI

The CLI brings your AI development team to the terminal and your CI pipelines. On macOS, install it with:

curl -fsSL https://downloads.boldcreate.ai/cli/install.sh | bash

On Windows, run this in PowerShell:

irm https://downloads.boldcreate.ai/cli/install.ps1 | iex

Then run boldcreate in any folder.

Work in an interactive session

Run boldcreate in a project folder and describe the problem. The agent reads the code, runs commands, edits files, and checks its work, and your organization’s skills load when the session starts.

Each step appears in the terminal as the agent works.

Pipe in a failing log

Pipe a failing log, diff, or stack trace into boldcreate -p -, and the agent reads the code, fixes the cause, and runs the tests. Print mode answers once and exits.

One command takes the test output and returns the fix.

Review commits in CI

Add --mode json to stream structured events your pipeline can parse, including every tool call and the final answer. The exit code tells the build what happened: 0 means passed, 1 failed, 2 a usage error, and 3 that the run is waiting for a person to approve.

Read the review in your pipeline with jq or any JSON tool.

Pick up where you left off

boldcreate sessions lists the sessions in the current folder, and boldcreate -c continues the most recent one with its context.

A resumed session keeps the earlier conversation and changes.

Enterprise Server

Enterprise Server moves to BoldCreate with a new layout and theme. Admins sign in at app.boldcreate.ai/dashboard, where the menu groups pages by job: Analytics, People, Cost Control, Models and Routing, and Settings. Names, emails, teams, and figures in these recordings are demo data.

Overview

The Overview shows how BoldCreate is used and how it performs for the period you pick: active users, total spend, adoption rate, agent sessions, requests, errors, and latency. A seat breakdown shows who uses BoldCreate, who stopped, and who never tried it.

Switch between Premium Model and Bring Your Own Key usage at the top.

Usage Explorer

Usage Explorer breaks spend, sessions, and tokens down by day, user, team, session, or model. Export the table, or schedule a report that’s emailed to the people you choose.

A scheduled report uses the team and user filters you set.

Users, teams, and budgets

Invite people by email with a team and a role, and see how many seats are in use. Set a budget for a team or a user, with a period, an optional automatic reset, and usage alerts. When someone needs more, they send a request, and approvers review it in Approval.

Approval's detail view shows each request next to its full context.

LLM keys, fallback, and settings

Add a provider key for the whole organization or for specific teams. A fallback policy retries failed requests on backup models in the order you set. Settings holds the audit log, single sign-on with OAuth or OpenID Connect, member permissions, and the budget approval workflow.

Keys stay masked in the list, and each one shows its scope.

Get started

Start with one task in your existing codebase. Start building, download the Desktop App, or request a demo to see how your team would assign a task, approve the spec, and review the pull request.

What's New in BoldCreate