diff --git a/content/guides/mistral-vibe-sandbox.md b/content/guides/mistral-vibe-sandbox.md
index 4816d43012..86dcf1ee1e 100644
--- a/content/guides/mistral-vibe-sandbox.md
+++ b/content/guides/mistral-vibe-sandbox.md
@@ -94,11 +94,11 @@ mistral-vibe/
 └── vibe.dockerfile
 ```
 
-Create `mistral-vibe/vibe.dockerfile`. The `shell-docker` template supplies
+Create `mistral-vibe/vibe.dockerfile`. The `shell-docker-image` base image supplies
 the agent account, Python, `uv`, and the sandbox's Docker tooling:
 
 ```dockerfile {title="mistral-vibe/vibe.dockerfile"}
-FROM docker/sandbox-templates:shell-docker
+FROM docker.io/sbx/shell-docker-image:latest
 
 ARG VIBE_VERSION
 
diff --git a/content/manuals/ai/sandboxes/customize/author/base-images.md b/content/manuals/ai/sandboxes/customize/author/base-images.md
index c4d9124a61..b9d546c0c4 100644
--- a/content/manuals/ai/sandboxes/customize/author/base-images.md
+++ b/content/manuals/ai/sandboxes/customize/author/base-images.md
@@ -22,7 +22,8 @@ To save and reuse an environment you've configured interactively, see
 
 ## Docker-provided images
 
-Docker's sandbox templates are published as
+To customize an existing agent, start from one of Docker's sandbox templates,
+published as
 `docker/sandbox-templates:<variant>`. They are based on Ubuntu and run as a
 non-root `agent` user with sudo access. Most variants include Git, Docker
 CLI, and common development tools like Node.js, Python, Go, and Java. Full
@@ -41,7 +42,22 @@ templates also include a C/C++ compiler toolchain for building native code.
 | `gemini`              | [Gemini CLI](https://github.com/google-gemini/gemini-cli)            |
 | `kiro`                | [Kiro](https://kiro.dev)                                             |
 | `opencode`            | [OpenCode](https://opencode.ai)                                      |
-| `shell`               | No agent pre-installed. Use for manual agent setup.                  |
+
+Docker rebuilds the `claude-code`, `claude-code-minimal`, `gemini`,
+`claude-code-docker`, and `gemini-docker` variants. The other variants in the
+table remain available as previously published images.
+
+### Build a custom agent on a shell image
+
+To package an agent that Docker doesn't provide, start from
+`docker.io/sbx/shell-image:latest`. This base image provides a development
+environment without a pre-installed agent, so you can add your agent's tooling
+on top. Use `docker.io/sbx/shell-docker-image:latest` if your workload needs
+Docker Engine inside the sandbox.
+
+If you need a different operating system or want to choose every package,
+[use your own Linux image](#use-your-own-linux-image). You'll need to prepare
+the tools and user account that Docker Sandboxes expects yourself.
 
 ## Use an image in a v3 workload
 
diff --git a/content/manuals/ai/sandboxes/customize/kits-v2.md b/content/manuals/ai/sandboxes/customize/kits-v2.md
index 6d83f41fa4..c19b0d470a 100644
--- a/content/manuals/ai/sandboxes/customize/kits-v2.md
+++ b/content/manuals/ai/sandboxes/customize/kits-v2.md
@@ -303,7 +303,7 @@ The agent's container image must provide:
 - HTTP proxy environment variables (`HTTP_PROXY`, `HTTPS_PROXY`, `NO_PROXY`) preserved across sudo.
 - The agent binary, either baked in or installed with [`setup.install`](#setup).
 
-Build on top of `docker/sandbox-templates:shell-docker` to get these base
+Build on top of `docker.io/sbx/shell-docker-image:latest` to get these base
 requirements.
 
 ## Agent instructions