Dropstone

Dropstone is a CLI-first AI coding assistant for developers and teams. It supports code generation, web search, memory across conversations, Remote MCP connections, and a companion dashboard.

AI编程助手 AI智能体开发 AI开发者工具编程代理

访问网站

What Dropstone is

Dropstone is a frontier-grade AI coding assistant built around a command-line workflow. The product pages present it as a way to code, refactor, test, and ship with higher weekly usage per dollar than other CLI tools, while keeping access to a terminal-first experience and a dashboard for related tasks.

The product is positioned for developers who want a coding agent that can generate code, search the web inside the agent, remember context across conversations, and connect external tools through Remote MCP. The download page also shows a self-contained CLI for macOS, Windows, and Linux, and the pricing page adds team and enterprise-oriented controls such as usage tiers, priority routing, and spend controls.

Core capabilities

Code generation with tiered models

Generate code directly in the agent, with model options that include Dropstone Fast, Pro, and Heavy depending on the plan.

Web search during the task

Use built-in web search inside the agent to gather outside context while working on a task.

Memory and reusable presets

Keep context across conversations and preserve useful state with saved agent presets and system prompts on higher plans.

Tool connection through Remote MCP

Connect external tools through Remote MCP, which the pricing page lists as a supported capability and connector surface.

Repo-aware workflow controls

Work in a repo-oriented workflow that supports code execution in a sandbox, file creation and editing, saved projects and workspaces, plan mode, and approval gates on tool calls.

Standalone CLI delivery

Install and run the CLI as a self-contained binary on macOS, Windows, and Linux, with no Node runtime or separate dependencies required.

Ways teams use Dropstone

Daily coding and refactoring
Generate new code, refactor existing files, or work through changes in a repo without leaving the terminal-centric workflow.
Reviewable agent-assisted changes
Use the planning and approval flow to stage edits before they are applied, which fits more controlled development work.
Research-backed implementation
Pull in outside context with web search inside the agent when a task depends on current documentation or external references.
Tool-connected workflows
Attach external services or internal tools through Remote MCP when a project needs the agent to work with more than the local repo.
Team and ongoing project work
Use the plan and dashboard surfaces to manage organized work across saved projects, workspaces, and higher-usage team plans.

Pros and Cons

Pros

CLI-first workflow with a companion dashboard.
Supports code generation, web search, memory across conversations, and Remote MCP tool connections.
Self-contained binary for macOS, Windows, and Linux.
Clear plan structure with Free, Pro, and Max options, plus team and enterprise billing controls.
Privacy and retention language is explicit, including no prompt, code, or output retention and no model training on customer data.

Cons

The public pages do not spell out every onboarding step or end-to-end workflow in detail.
Some capabilities are described broadly on the pricing page, so readers may need the docs for implementation specifics and integration depth.

FAQ

What is Dropstone?

Dropstone is an AI coding assistant for terminal and workspace-based development. The source pages describe it as a CLI-first product with a dashboard, model choices, and tools for code generation, web search inside the agent, memory across conversations, and Remote MCP connections.

Which platforms does Dropstone support?

The download page says the CLI is available for macOS, Windows, and Linux. It also lists a self-contained binary with support for macOS 11+, Windows 10+, and modern Linux on arm64 and x86_64.

Does Dropstone have both free and paid plans?

The pricing page shows a Free plan and paid Pro and Max plans. It also presents team and enterprise-oriented controls such as organizational spend controls, multi-currency support, and invoicing or net terms for enterprise billing.

How does Dropstone handle setup and data handling?

The download page describes the CLI as a self-contained, statically linked binary with no Node runtime or separate dependencies. The pricing page also states that prompt, code, and output are not retained, and that model training on your data is strictly disabled.

What kinds of tasks can I do with it?

The source pages do not describe every workflow detail, but they do show supported actions such as generating code, editing files in a repo, planning changes before applying them, and using approval gates on tool calls.

Quick Facts

Category: AI coding assistant
Primary interface: Command-line tool with dashboard support
Platforms: macOS, Windows, Linux
Distribution: Self-contained binary
Source domain: dropstone.io
Typical users: Developers and teams