HolyClaude: A Docker-Packaged AI Workstation for Claude Code and Eight AI CLIs
AI coding workstation: Claude Code + web UI + 8 AI CLIs + headless browser + 50+ tools.
At a glance
- What is it?
- HolyClaude bundles Claude Code, a web UI, a headless browser, and 50-plus developer tools into one Docker image. The pitch is simple: stop configuring, start building. The reality is a heavy, opinionated container that trades flexibility for convenience.
- Who is it for?
- Adopt HolyClaude if you want a pre-assembled, containerized AI coding environment and you accept its fixed version pins and CloudCLI account requirement. Skip it if you prefer to manage your own Node, Playwright, and AI CLI versions, or if you cannot tolerate a mandatory third-party account for the web UI.
- Can I use it commercially?
- Yes. MIT is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.
- Is it still maintained?
- Yes. The repository received new commits within the last day.
- What is it written in?
- Mainly JavaScript, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on November 15, 2026, and from our analysis. They are not legal advice.
Star history
- CoderLuii/HolyClaude2,662 ★ · +195
| Repository | 2026-07-31 | 2026-11-15 |
|---|---|---|
| CoderLuii/HolyClaude | 2.5k | 2.7k |
195 Stars in the last 107 days, measured from the daily GitHub snapshot. Compare with other projects
Editorial analysis
The Problem It Solves: Docker Pain, Not AI Pain
HolyClaude exists because running Claude Code inside Docker is notoriously fiddly. The README lists the usual failure cascade: Chromium won't launch because Docker's shared memory is 64MB, Xvfb isn't configured, the container UID mismatches the host, Claude Code's installer hangs on a root-owned WORKDIR, and SQLite locks on NAS mounts. HolyClaude is the author's accumulated fix for those problems, packaged as an image. The target user is someone who wants Claude Code plus a browser UI plus a headless browser plus a stack of AI CLIs without spending hours wiring them together. It is not for people who want a minimal, single-purpose container. It is for people who want a full workstation and are willing to accept a large, pre-configured image.
What the Image Actually Contains
The README describes a container with Claude Code, a web UI, a headless browser, eight AI CLIs, and 50-plus developer tools. The v1.5.7 release notes give concrete version numbers: Node 26.7.0, Playwright 1.62.0, tsx 4.23.12, pnpm 11.21.0, Vite 8.2.1, esbuild 0.28.2, ESLint 10.8.1. npm stays at 11.19.0 because npm 12 rejects CloudCLI's verified shrinkwrap. TypeScript stays at 6.0.3 because the 7.0.2 native binary carries fixed-version Go findings. CloudCLI is pinned at 1.36.3 with patched nested security dependencies. These pins are not incidental. They are deliberate choices made to keep a complex dependency graph working. The trade-off is obvious: you get a working system, but you do not get the latest versions of anything unless the maintainer updates the image. The release workflow checks a contract file against the Dockerfile and Compose files before building images, which is a serious attempt to keep the bundle coherent.
How It Works: Authentication and the Web UI
HolyClaude runs the real Claude Code CLI, not a wrapper or proxy. Authentication works through the web UI using OAuth, same as desktop Claude Code, for Max or Pro plan users. API key users can set the key through the web UI. The README is explicit that HolyClaude operates no credential relay. Bundled tools read credentials from container files, bind mounts, or environment variables, then contact providers directly. The web UI requires creating a CloudCLI account before you can sign in with your Anthropic account. That is a third-party account you might not want. The README says no extra cost, but it does not explain what CloudCLI is or why it is required. That is a genuine gap in the documentation. You are trusting an intermediary for the login flow, even if credentials are not relayed.
Getting Started: Commands and Configuration
The quick start is short. Create a folder, copy a docker-compose template, run docker compose up -d, open http://localhost:3001, create a CloudCLI account, sign in with Anthropic. The README stresses that there are no .env files and no pre-configuration. Two compose templates exist: a minimal quick template and a full template with all options documented. Environment variables are listed in a dedicated section, but the truncated material does not show their names. The image is available on Docker Hub as coderluii/holyclaude. For remote access, the README explicitly warns against port-forwarding and points to a section on remote access and exposure.
The Version Pin Dilemma
The v1.5.7 release notes show a maintenance pattern that will matter to adopters. Node, Playwright, tsx, pnpm, Vite, esbuild, and ESLint all move forward. npm and TypeScript stay pinned because newer versions break CloudCLI or carry security findings. This is a classic dependency hell situation, and HolyClaude solves it by freezing versions. The cost is that you cannot upgrade individual tools without waiting for a new HolyClaude release. The release notes also mention downstream backports for Netlify image parsing, FFmpeg, and Azure CLI's cryptography runtime. That level of patching is impressive, but it also means the image is a moving target. The maintainer has done the work, but you are tied to their schedule. If a critical security fix lands in one of the pinned tools, you depend on the project's release cadence, not the upstream project's.
A Real Limitation: You Must Trust the Bundle
HolyClaude is a black box in the sense that you do not assemble it yourself. The README lists many tools, but the truncated material does not enumerate all 50-plus tools or the eight AI CLIs. You cannot verify what is inside without pulling the image and inspecting it. The README's promise that every edge case has been handled is a claim, not a fact you can check from the documentation. The known issues section exists. This is the wrong tool if you need a minimal base image, if you need to audit every dependency, or if you need to run different versions of Node or Playwright for different projects. The container is a full workstation, not a component you slot into an existing setup.
Alternatives: Roll Your Own or Use a Thinner Base
The obvious alternative is to build your own Dockerfile. You would start from a Node image, install Claude Code manually, add Playwright and Chromium, configure Xvfb, and set the shared memory size. That is exactly the two-hour setup HolyClaude claims to save. The difference in approach is control versus convenience. With your own Dockerfile, you choose every version and you understand every layer. With HolyClaude, you accept someone else's choices. Another alternative is to skip Docker entirely and run Claude Code directly on your host, using a browser-based IDE like VS Code with the Claude extension. That avoids the container complexity but loses the portability and the headless browser. The README itself has an Alternatives section, but the content is truncated. The point is that HolyClaude's value is the pre-solved Docker problems, and that value only matters if you actually need a containerized environment.
Maintenance, Upgrades, and Licensing
The project is under the MIT license, which means you can use, modify, and redistribute it freely, with attribution. The README has an Upgrading section. The release workflow includes rollback evidence and retries only a failed browser snapshot once, which suggests a mature release process. The contract file that the workflow checks against the Dockerfile and Compose files is a notable engineering practice. It means the image and the compose files are verified to match before a build. For maintenance cost, expect to track releases. The version pinning means you will want to pull new images regularly to get updated tools. The README's claim that you can use your existing Claude Code subscription means no additional AI cost, but the CloudCLI account is an unavoidable dependency. The project is actively maintained, with three releases in August 2026 alone, so the upgrade path is alive.
Editorial conclusion
Adopt HolyClaude if you want a pre-assembled, containerized AI coding environment and you accept its fixed version pins and CloudCLI account requirement. Skip it if you prefer to manage your own Node, Playwright, and AI CLI versions, or if you cannot tolerate a mandatory third-party account for the web UI. Before adopting, verify the current release notes for version pin changes, confirm that your Anthropic subscription works with the bundled OAuth flow, and test the container's UID mapping against your host's permissions.
Frequently asked questions
What is HolyClaude?
A containerized AI development workstation that bundles the Claude Code CLI, a web interface called CloudCLI, a headless browser with Playwright configured, eight AI CLIs and more than 50 developer tools, so the environment can be reached from a browser instead of a locally installed toolchain.
How do I start the HolyClaude container?
Create a directory, write a docker-compose.yaml from one of the two templates, and run docker compose up. The quick template publishes the web UI on 127.0.0.1:3001, sets shm_size to 2g, and mounts ./data/claude and ./workspace.
Does HolyClaude need a separate Claude subscription?
No. It works with a Claude Max or Pro plan authenticated through the web interface using OAuth, or with an Anthropic API key entered the same way, and the README states it runs the real Claude Code CLI rather than a wrapper or a proxy.
Which security findings does the current HolyClaude release accept?
Version 1.6.4 carries an acceptance through 2026-10-04 for the exact known Critical and High findings recorded for the native Full and Slim images on AMD64 and ARM64, including findings in HolyClaude's custom FFmpeg build and bundled tool dependencies. Any new or changed finding still blocks publication.
Which CPU architectures can HolyClaude be built for?
amd64 and arm64 only. The build script exits with an Unsupported TARGETARCH message for anything else, and the pinned Debian Chromium build covers the same two architectures.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/coderluii-holyclaude)Star history chart for your README
[](https://hysenlabs.com/projects/coderluii-holyclaude)Was this useful?