BoldCreate CLI: Bring Agentic Development to Your Terminal

Bring AI Into Your Terminal Workflow

For developers, the terminal is more than a place to run commands. It is where builds run, tests fail, logs reveal problems, and development workflows come together.

Switching to another tool to explain a build failure or provide project context can interrupt that flow.

BoldCreate CLI brings agentic development directly into your terminal. Describe a task, give the agent relevant context, and let it help investigate issues, modify code, and run commands within your development workflow.

Whether you’re working locally or using terminal-based automation, the CLI gives you another way to work with the BoldCreate agentic development platform.

Start a BoldCreate Session

Open your terminal in the project folder and run:

boldcreate

You can also provide an initial request:

boldcreate "Review the authentication flow"

BoldCreate starts a session in your current workspace. You can describe a task immediately or continue the conversation interactively as the agent works.

Give the Agent More Context

The right context helps an agent understand what you want to accomplish.

BoldCreate CLI lets you include text files directly in your initial request using @path arguments. For example:

boldcreate "Review this changelog" @CHANGELOG.md

You can also provide supported image files, including screenshots, diagrams, and mockups, by specifying their paths. This helps you share error messages or visual references alongside your instructions.

Resume Previous Conversations

You don’t have to start from scratch every time. BoldCreate saves sessions locally so you can return to previous conversations and continue your work.

List saved sessions:

boldcreate sessions

Continue the most recent session in the current folder:

boldcreate -c

Open the session picker:

boldcreate -r

You can also resume a specific session using boldcreate --session ID or name a session to make it easier to find later.

Sessions are associated with the folder where they were started. If you cannot find a session, return to the original project folder or use boldcreate sessions --all to list sessions across folders.

Connect External Tools Through MCP

Development tasks can require information and capabilities beyond your local codebase.

BoldCreate CLI supports the Model Context Protocol (MCP), which lets you connect external tools and services to the agent.

For example, you can add a Playwright MCP server with:

boldcreate mcp add playwright -- npx -y @playwright/mcp@latest --headless

List configured servers and check their connection status:

boldcreate mcp list

Inside an interactive session, use /mcp to manage servers, inspect their tools, and handle connection or sign-in requirements.

MCP integrations extend the tools available to the agent, depending on the servers you configure and enable.

Manage Long-Running Tasks

Builds, test suites, development servers, and other commands can take time to finish. BoldCreate CLI can run long-running commands in the background, allowing their progress and output to be checked without unnecessarily restarting the work.

Use:

/tasks

to view and manage background tasks in an interactive session. You can inspect task status and output, and stop running tasks when needed.

Background tasks continue while BoldCreate is running, but they stop when you exit the CLI. Keep this in mind when running development servers or other commands that need to remain active.

Configure How the Agent Works

Different tasks call for different levels of planning, reasoning, and oversight. BoldCreate CLI lets you configure the model, reasoning effort, agent mode, and approval behavior.

For example, switch to planning mode with:

/agent plan

Set the approval mode to Manual with:

/mode manual

You can also use /agent agent for implementation work or /agent ask to ask questions without making changes. Supported reasoning levels include none, minimal, low, medium, high, and xhigh, depending on the selected model.

Choose settings that suit the task and your preferred level of control. In Manual approval mode, the agent requests authorization for actions that require approval before proceeding.

Bring Agentic Development to Your Terminal

Your terminal already provides access to your project files, development commands, and working context. BoldCreate CLI brings an AI agent into that environment, helping you investigate problems, work through development tasks, connect external tools, and return to previous conversations.

Start a session. Provide context. Delegate a task. Review the result.

Ready to get started?

Explore BoldCreate CLI and bring agentic development to your terminal and CI workflows.

Frequently Asked Questions