Introducing Agent Context Explorer (ACE) for Cursor / VS Code

This project grew organically as I learned more and more aspects of AI agent configurations. I first learned about rules, then commands, and later skills. The name and the focus have changed along the way, until, after several iterations, it became the Agent Context Explorer (ACE).

While working on a poly-repository project, I found myself wanting a single place to see agents configurations without traversing through directories. An early incarnation included optimistic context determination and an export tool to share context with agents directly.

ACE is a VS Code extension that surfaces your AI agent artifacts in one tree. The Workspaces view shows Cursor and Claude Code rules, commands, skills, and project specs for every open project.

Workspaces View

The Workspaces view is intended to display related projects and the Claude Code or Cursor artifacts therein.

Agents View

The Agents view shows your global configuration across ~/.cursor/, ~/.claude/, and ~/.agents/ — the commands, skills, and agent definitions available to your AI tools at the user level.

MCP

The early export implementation was later dropped in favour of an MCP server. In Cursor it registers automatically. For other tools, a small addition to your MCP configuration is all it takes.

ACE is viewer-only by design. It scans and displays the intentional artifacts you've already created — it doesn't manage them. The idea is visibility, not control.

Install from Open VSX (https://open-vsx.org/extension/fancy-bread/agent-context-explorer) for Cursor or the Visual Studio Marketplace (https://marketplace.visualstudio.com/items?itemName=fancy-bread.agent-context-explorer) for VS Code, then open the ACE icon in your sidebar.

The tooling around AI-assisted development is still taking shape. Each new capability opens up questions about what else is possible. ACE is one answer to a question I kept running into.