briq gives coding agents real databases, sandboxes to run code in and whole apps in one MCP call: isolated, EU-hosted, billed by the second, with a capped budget.
Developer toolsTypeScript · JavaScriptGitHub @briq-cloudOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Run /nomorevibe verify in your project folder to verify ownership and manage the introduction yourself.
Verify in one minuteAsk about other verification options
The image behind briq's agent-home recipe: the vendors' agent CLIs, unmodified and pinned, on the briq sandbox
Taken from the repository README
Loading the full README.
Read the full README ↗The image behind briq_run({ recipe: "agent-home" }) on briq: one person's
computer in the cloud, where they sign in to each vendor's agent CLI with the vendor's own flow,
and their agents run on their own plans.
ghcr.io/briq-cloud/agent-home:latest
The briq sandbox (Ubuntu 24.04, Python, Node.js, git, the usual tools) plus:
| CLI | Command | Sign-in |
|---|---|---|
| Claude Code | claude | claude auth login |
| Codex | codex | codex login --device-auth |
| Copilot CLI | copilot | copilot login --device-code |
| Cursor | agent | NO_OPEN_BROWSER=1 agent login |
| Grok Build | grok | grok login --device-auth |
| Mistral Vibe | vibe | vibe --setup |
| Gemini CLI | gemini | a key only (GEMINI_API_KEY), enforced by its system settings: it serves no consumer plan since June 2026 |
Other providers' keys (the GLM, Kimi, MiniMax and Alibaba coding plans, DeepSeek, OpenRouter) run on Claude Code or Codex as profiles, added on the runner's page.
and ttyd for a terminal page, and agent-home-runner: it drives the CLIs over ACP for apps and
keeps their sessions while nobody is connected (runner/README.md). Every version, as the last build left them:
REFERENCE.md. Inside a home, agent-home-versions prints the same.
2026-10-27) never moves. latest follows the newest build, every Monday and on every change
here, for linux/amd64.DISABLE_AUTOUPDATER and DISABLE_UPDATES (Claude Code),
/etc/codex/config.toml, COPILOT_AUTO_UPDATE=false, GROK_DISABLE_AUTOUPDATER and
/etc/grok/managed_config.toml, /etc/gemini-cli/settings.json. Cursor and Vibe read theirs
from the user's own config. Gemini's usage statistics are off in its system settings.SHA256SUMS, PyPI's); Cursor and Grok publish none, so their
hashes are recorded. Every sha256 and the base image's digest are in REFERENCE.md.gh attestation verify oci://ghcr.io/briq-cloud/agent-home:<tag> --repo briq-cloud/agent-home.xdg-open or $BROWSER; here both
are show-url, which shows the link as plain text in a terminal window of its own.Sign-ins live where each CLI's variable says (CLAUDE_CONFIG_DIR, CODEX_HOME, COPILOT_HOME,
GROK_HOME, VIBE_HOME, XDG_CONFIG_HOME). On briq they are on the home's volume, which is never
forked or backed up.
docker run --rm -it ghcr.io/briq-cloud/agent-home:latest bash
docker build -t agent-home .
docker run --rm agent-home agent-home-versions
MIT licensed. The briq service is operated by Uplab SAS under its terms.
Small ACP benchmarks: free mock checks and explicit, budgeted provider-key experiments.