Skip to content
| Marketplace
Sign in
Visual Studio Code>Visualization>Agent CodeWalkNew to Visual Studio Code? Get it now.
Agent CodeWalk

Agent CodeWalk

agent-codewalk

|
1 install
| (0) | Free
Follow guided, local walkthroughs of code changed or explained by Codex, Claude Code, and OpenCode.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

Agent CodeWalk logo

Agent CodeWalk

Turn an AI agent's work into a guided tour of the code—one meaningful block at a time.

English · 简体中文 · GitHub

Agent CodeWalk product preview

Agent CodeWalk lets you read what Codex, Claude Code, or OpenCode changed—or what it explained—as a walkthrough inside VS Code, Cursor, VSCodium, and compatible desktop editors. Every step opens the relevant file, highlights the exact block, and keeps the agent's reasoning beside the code instead of leaving you to hunt through a chat transcript.

No extra model, cloud service, or API key is required. The agent you already use creates the explanation, and the baseline and walkthrough data stay on your machine.

What you get

  • A dependency graph for following the order code runs, plus a complete file-grouped view
  • Nested high-level steps that open into focused implementation detail
  • Theme-aware highlights for added, modified, deleted, renamed, and contextual code
  • A before/after comparison for steps that replaced existing text
  • CodeLens, status bar, keyboard, sidebar, and searchable step navigation
  • Safe relocation when code moves, with an explicit stale state when the target is ambiguous
  • English and Simplified Chinese interface text that follows the editor language

Install

The easiest update path is a registry installation:

  • Visual Studio Marketplace
  • Open VSX Registry

You can also install from a terminal:

# VS Code
code --install-extension agent-codewalk.agent-codewalk

# Cursor
cursor --install-extension agent-codewalk.agent-codewalk

# VSCodium
codium --install-extension agent-codewalk.agent-codewalk

Offline and pinned packages for Linux x64, Windows x64, macOS Intel, and macOS Apple silicon are attached to every GitHub release, together with SHA-256 checksums.

One-time agent setup

  1. Run Agent CodeWalk: Setup Agent Integrations from the Command Palette.
  2. Review the exact user-level files shown in the confirmation and choose Install.
  3. Restart active Codex, Claude Code, or OpenCode sessions so they load the local MCP companion and skill.
  4. Run Agent CodeWalk: Diagnose Installation if a walkthrough does not appear. The command asks each agent which MCP servers it actually loaded.

Setup creates backups, rolls back a failed agent integration, and refuses to overwrite a skill or MCP entry it does not own. Agent CodeWalk: Remove Agent Integrations removes only owned resources and keeps published walkthroughs.

Use it naturally

Ask your agent to change code as usual:

Add retry handling to the upload path and cover the failure cases.

The agent records a baseline before its first edit and publishes a walkthrough after it verifies the task. The local companion checks the real diff and requires every text hunk to be covered.

Or ask it to explain existing code without making a change:

Trace how a published walkthrough reaches the sidebar.

An explanation walkthrough needs no baseline, but every step must still point at code that exists now.

Navigate

Open Agent CodeWalk from the container switcher at the top of the right-hand Secondary Side Bar, accept the publication notification, or run Agent CodeWalk: Open Latest Walkthrough. VS Code uses this right-side location for a fresh layout and remembers any location you choose later.

Action Shortcut
Next visible step Alt+]
Previous visible step Alt+[
Switch graph/file view Alt+\
Jump to any step Ctrl+Alt+W (Cmd+Alt+W on macOS)

The current step also appears as a CodeLens above its block and as progress in the status bar. Select Compare with before when a modified step has baseline text available.

Local-first and private

Agent CodeWalk includes no telemetry, makes no network requests, and does not call a model API. The Rust companion communicates over standard input/output and never listens on a port. Sessions keep paths, explanations, ranges, hashes, and a bounded amount of replaced text for step comparison; they do not copy whole source files.

Read the complete privacy and security model, architecture, and troubleshooting guide in the project repository.

Requirements

  • Desktop VS Code 1.106 or newer, or a compatible Cursor/VSCodium build with native Secondary Side Bar view containers
  • Codex, Claude Code, or OpenCode installed before running integration setup
  • A platform package matching the machine that runs the extension host

Browser-hosted editors cannot start the local companion. Remote SSH and WSL are supported by installing the matching package in the workspace extension host.

Agent CodeWalk is open source under the MIT License.

  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
© 2026 Microsoft