Skip to content

🤖 Open-source, local-first alternative to Grok Bot, OpenAI dots and Claude Duet. Your own AI bots on any model: GPT, Claude, Gemini, Grok, DeepSeek, Kimi, GLM and 70+ more through one Modelflare key. Reach them in the app, WeChat or Feishu; they remember you, work in the background, run commands and use your computer. Pay per use.

ci(docker): 发版时把 x64 + arm64 镜像发布到 GHCR,README 改成一条 docker run

dyk987ecee7b0349c
diff --git a/.github/workflows/docker.yml b/.github/workflows/docker.yml
new file mode 100644
index 0000000..44785a2
--- /dev/null
+++ b/.github/workflows/docker.yml
@@ -0,0 +1,208 @@
+# Docker: publish the app image to GitHub Container Registry (GHCR) on every release.
+#
+# How to use
+# - Release: pushing a tag `v<version>` (the same tag that starts the Release workflow) builds the
+# image for linux/amd64 and linux/arm64 and publishes it as
+# ghcr.io/modelflare-ai/anymodel-bot:<version> every release
+# ghcr.io/modelflare-ai/anymodel-bot:latest only releases without "-" (not v0.2.0-rc.1)
+# Users then need one command (the README and the website show it):
+# docker run -d --name amb -p 3600:3600 -v amb-data:/data ghcr.io/modelflare-ai/anymodel-bot
+# - Publish an existing tag (one created before this workflow existed): Actions -> Docker ->
+# "Run workflow" on master with `tag` set, e.g. v0.1.10. That tag is checked out and published.
+# - Test build: "Run workflow" with `tag` empty, on any branch. Both architectures are built from
+# that branch, nothing is pushed.
+#
+# One-time setup after the first publish: GHCR creates the package as private. Open
+# github.com/orgs/modelflare-ai/packages -> anymodel-bot -> Package settings -> Change visibility ->
+# Public, or `docker pull` asks users to log in.
+#
+# Each architecture builds on a runner of its own CPU (the image installs per-arch native packages
+# in dsh-runtime, like the installers in release.yml) and pushes by digest; the merge job then joins
+# the two digests into one multi-arch tag.
+
+name: Docker
+
+on:
+ push:
+ tags: ["v*"]
+ workflow_dispatch:
+ inputs:
+ tag:
+ description: "Release tag to publish (e.g. v0.1.10). Empty: test build of this branch, nothing is pushed."
+ type: string
+ default: ""
+
+permissions:
+ contents: read
+
+concurrency:
+ group: docker-${{ inputs.tag || github.ref }}
+ cancel-in-progress: ${{ github.ref_type != 'tag' && inputs.tag == '' }}
+
+env:
+ # GHCR image names must be lowercase.
+ IMAGE: ghcr.io/modelflare-ai/anymodel-bot
+
+jobs:
+ meta:
+ name: Version check
+ runs-on: ubuntu-24.04
+ timeout-minutes: 5
+ outputs:
+ ref: ${{ steps.version.outputs.ref }}
+ version: ${{ steps.version.outputs.version }}
+ push: ${{ steps.version.outputs.push }}
+ latest: ${{ steps.version.outputs.latest }}
+ steps:
+ - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+ with:
+ persist-credentials: false
+ ref: ${{ inputs.tag || github.ref }}
+ sparse-checkout: package.json
+ sparse-checkout-cone-mode: false
+
+ - name: Read package.json version and match it against the tag
+ id: version
+ env:
+ INPUT_TAG: ${{ inputs.tag }}
+ REF_TYPE: ${{ github.ref_type }}
+ REF_NAME: ${{ github.ref_name }}
+ run: |
+ version=$(node -p 'require("./package.json").version')
+ tag=""
+ if [ -n "$INPUT_TAG" ]; then tag=$INPUT_TAG; elif [ "$REF_TYPE" = tag ]; then tag=$REF_NAME; fi
+ if [ -n "$tag" ] && [ "$tag" != "v$version" ]; then
+ echo "::error::Tag $tag does not match package.json version $version (expected tag v$version)"
+ exit 1
+ fi
+ if [ -n "$tag" ]; then
+ echo "ref=refs/tags/$tag" >> "$GITHUB_OUTPUT"
+ echo "push=true" >> "$GITHUB_OUTPUT"
+ if [[ "$tag" == *-* ]]; then echo "latest=false" >> "$GITHUB_OUTPUT"; else echo "latest=true" >> "$GITHUB_OUTPUT"; fi
+ echo "Publishing $IMAGE:$version"
+ else
+ echo "ref=$GITHUB_REF" >> "$GITHUB_OUTPUT"
+ echo "push=false" >> "$GITHUB_OUTPUT"
+ echo "latest=false" >> "$GITHUB_OUTPUT"
+ echo "Test build of $GITHUB_REF ($version), nothing is pushed"
+ fi
+ echo "version=$version" >> "$GITHUB_OUTPUT"
+
+ build:
+ name: linux/${{ matrix.arch }}
+ needs: meta
+ runs-on: ${{ matrix.runner }}
+ timeout-minutes: 45
+ permissions:
+ contents: read
+ packages: write
+ strategy:
+ fail-fast: true
+ matrix:
+ include:
+ - { arch: amd64, runner: ubuntu-24.04 }
+ - { arch: arm64, runner: ubuntu-24.04-arm }
+ steps:
+ - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+ with:
+ persist-credentials: false
+ ref: ${{ needs.meta.outputs.ref }}
+
+ # The Dockerfile compiles the bundled gateway from this submodule.
+ - name: Fetch magpie
+ run: git submodule update --init --depth 1 vendor/magpie
+
+ - uses: docker/setup-buildx-action@f87e5991a6d7451dcb8d9637bfbc97413f497069 # v4.4.1
+
+ - uses: docker/login-action@81c6f93d4bc991e510b342a00b419b3bb4efc5c6 # v4.7.0
+ if: needs.meta.outputs.push == 'true'
+ with:
+ registry: ghcr.io
+ username: ${{ github.actor }}
+ password: ${{ github.token }}
+
+ - name: Build${{ needs.meta.outputs.push == 'true' && ' and push by digest' || ' (test, not pushed)' }}
+ id: build
+ uses: docker/build-push-action@c3c9e263c25d99ce0380d002d59b67737d91b0dc # v7.4.0
+ with:
+ context: .
+ platforms: linux/${{ matrix.arch }}
+ tags: ${{ env.IMAGE }}
+ outputs: ${{ needs.meta.outputs.push == 'true' && 'type=image,push-by-digest=true,name-canonical=true,push=true' || 'type=cacheonly' }}
+ # No attestation manifests: registry UIs and NAS apps list them as an extra "unknown/unknown" platform.
+ provenance: false
+ sbom: false
+ labels: |
+ org.opencontainers.image.title=AnyModelBot
+ org.opencontainers.image.description=Local-first AI bots with any model, pay as you go
+ org.opencontainers.image.source=https://github.com/${{ github.repository }}
+ org.opencontainers.image.url=https://anymodel.bot/
+ org.opencontainers.image.version=${{ needs.meta.outputs.version }}
+ org.opencontainers.image.revision=${{ github.sha }}
+ org.opencontainers.image.licenses=MIT
+ cache-from: type=gha,scope=docker-${{ matrix.arch }}
+ cache-to: type=gha,mode=max,scope=docker-${{ matrix.arch }}
+
+ - name: Export digest
+ if: needs.meta.outputs.push == 'true'
+ env:
+ DIGEST: ${{ steps.build.outputs.digest }}
+ run: |
+ mkdir -p "$RUNNER_TEMP/digests"
+ touch "$RUNNER_TEMP/digests/${DIGEST#sha256:}"
+
+ - uses: actions/upload-artifact@cf430e030ddbb5b0abf93d22962f4752f3646cd9 # v7.0.2
+ if: needs.meta.outputs.push == 'true'
+ with:
+ name: docker-digest-${{ matrix.arch }}
+ path: ${{ runner.temp }}/digests/*
+ if-no-files-found: error
+ overwrite: true
+ retention-days: 1
+
+ merge:
+ name: Publish multi-arch tags
+ needs: [meta, build]
+ if: needs.meta.outputs.push == 'true'
+ runs-on: ubuntu-24.04
+ timeout-minutes: 10
+ permissions:
+ contents: read
+ packages: write
+ steps:
+ - uses: actions/download-artifact@9000827ccba6bdab643e8b6fd33ac0654aef8333 # v8.0.2
+ with:
+ pattern: docker-digest-*
+ path: ${{ runner.temp }}/digests
+ merge-multiple: true
+
+ - uses: docker/setup-buildx-action@f87e5991a6d7451dcb8d9637bfbc97413f497069 # v4.4.1
+
+ - uses: docker/login-action@81c6f93d4bc991e510b342a00b419b3bb4efc5c6 # v4.7.0
+ with:
+ registry: ghcr.io
+ username: ${{ github.actor }}
+ password: ${{ github.token }}
+
+ - name: Create the multi-arch tags
+ working-directory: ${{ runner.temp }}/digests
+ env:
+ VERSION: ${{ needs.meta.outputs.version }}
+ LATEST: ${{ needs.meta.outputs.latest }}
+ REPO_URL: https://github.com/${{ github.repository }}
+ run: |
+ tags=(-t "$IMAGE:$VERSION")
+ if [ "$LATEST" = true ]; then tags+=(-t "$IMAGE:latest"); fi
+ sources=()
+ for d in *; do sources+=("$IMAGE@sha256:$d"); done
+ if [ "${#sources[@]}" -ne 2 ]; then
+ echo "::error::Expected 2 digests (amd64, arm64), found ${#sources[@]}"
+ exit 1
+ fi
+ # Index annotations: GHCR reads the description and the source repository of a multi-arch image from here.
+ docker buildx imagetools create "${tags[@]}" \
+ --annotation "index:org.opencontainers.image.source=$REPO_URL" \
+ --annotation "index:org.opencontainers.image.description=Local-first AI bots with any model, pay as you go" \
+ --annotation "index:org.opencontainers.image.licenses=MIT" \
+ "${sources[@]}"
+ docker buildx imagetools inspect "$IMAGE:$VERSION"
diff --git a/Dockerfile b/Dockerfile
index 1135995..e92ecb2 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,5 +1,6 @@
# anymodel bot:一个镜像同时包含机器人运行时 + 内置 magpie 模型网关
-# 构建前需检出子模块:git submodule update --init --depth 1 vendor/magpie
+# 发版时由 .github/workflows/docker.yml 发布到 ghcr.io/modelflare-ai/anymodel-bot(x64 + arm64),用户直接 docker run 那个镜像。
+# 自己构建前需检出子模块:git submodule update --init --depth 1 vendor/magpie
# docker build -t anymodel-bot .
# docker run -d --name amb -p 3600:3600 -v amb-data:/data anymodel-bot
# docker exec amb cat /data/ui-token # 首次打开 http://<主机>:3600/?token=<令牌>
diff --git a/README.md b/README.md
index 28d7cf9..2e6d8de 100644
--- a/README.md
+++ b/README.md
@@ -112,7 +112,16 @@ sha256sum -c SHA256SUMS --ignore-missing
macOS 用 `shasum -a 256 <文件名>`,Windows PowerShell 用 `Get-FileHash <文件名>`,再和 `SHA256SUMS` 里的值对照。
-**从源码运行,或用 Docker**
+**用 Docker 部署到 NAS 或服务器**(镜像同时支持 x64 和 arm64)
+
+```bash
+docker run -d --name amb -p 3600:3600 -v amb-data:/data ghcr.io/modelflare-ai/anymodel-bot
+docker exec amb cat /data/ui-token # 然后打开 http://<主机>:3600/?token=<令牌>
+```
+
+不要把 3600 端口直接暴露到公网。
+
+**从源码运行**
```bash
git clone https://github.com/modelflare-ai/AnyModelBot.git
@@ -122,14 +131,7 @@ npm run setup # 拉取并编译内置网关
npm start # 本地管理界面:http://127.0.0.1:3600/
```
-```bash
-git submodule update --init --depth 1 vendor/magpie # 镜像要从源码编译内置网关
-docker build -t anymodel-bot .
-docker run -d --name amb -p 3600:3600 -v amb-data:/data anymodel-bot
-docker exec amb cat /data/ui-token # 然后打开 http://<主机>:3600/?token=<令牌>
-```
-
-从源码构建需要 Node.js 22.13+ 和 Go 1.26.8+。更多内容见[开发文档](./docs/development.md)。
+从源码构建需要 Node.js 22.13+ 和 Go 1.26.8+。想自己构建 Docker 镜像,见[开发文档](./docs/development.md)。
# 🌠 截图
diff --git a/README_en.md b/README_en.md
index bf757e6..61499bf 100644
--- a/README_en.md
+++ b/README_en.md
@@ -112,7 +112,16 @@ sha256sum -c SHA256SUMS --ignore-missing
On macOS use `shasum -a 256 <file>`; in Windows PowerShell use `Get-FileHash <file>`. Compare the result with the line in `SHA256SUMS`.
-**Run from source or in Docker**
+**Run it in Docker on a NAS or server** (the image covers both x64 and arm64)
+
+```bash
+docker run -d --name amb -p 3600:3600 -v amb-data:/data ghcr.io/modelflare-ai/anymodel-bot
+docker exec amb cat /data/ui-token # then open http://<host>:3600/?token=<token>
+```
+
+Don't expose port 3600 directly to the internet.
+
+**Run from source**
```bash
git clone https://github.com/modelflare-ai/AnyModelBot.git
@@ -122,14 +131,7 @@ npm run setup # fetches and builds the bundled gateway
npm start # opens the local console at http://127.0.0.1:3600/
```
-```bash
-git submodule update --init --depth 1 vendor/magpie # the image builds the gateway from source
-docker build -t anymodel-bot .
-docker run -d --name amb -p 3600:3600 -v amb-data:/data anymodel-bot
-docker exec amb cat /data/ui-token # then open http://<host>:3600/?token=<token>
-```
-
-Requirements for building from source: Node.js 22.13+ and Go 1.26.8+. More in the [development guide](./docs/development_en.md).
+Requirements for building from source: Node.js 22.13+ and Go 1.26.8+. To build the Docker image yourself, see the [development guide](./docs/development_en.md).
# 🌠 Screenshots
diff --git a/docs/development.md b/docs/development.md
index 920bc11..7620dc5 100644
--- a/docs/development.md
+++ b/docs/development.md
@@ -66,10 +66,12 @@ npm run test:dsh-bundle # 用 Electron 跑真实 DSH 集成测试,加载打
- 界面只监听 127.0.0.1,且每次启动生成随机访问令牌,其他本机程序无法直接调用管理接口。
- 退出时先停渠道、再停 magpie;若上次被强杀残留 magpie 进程,下次启动会自动清理(Linux/macOS)。
-发布:改 `package.json` 的版本号并提交,推送 `v<版本>` 标签,`.github/workflows/release.yml` 会在六种系统 / 架构上各自打包,并把安装包和 `SHA256SUMS` 发到 GitHub Release。签名与公证所需的仓库密钥见该文件开头的注释。
+发布:改 `package.json` 的版本号并提交,推送 `v<版本>` 标签,`.github/workflows/release.yml` 会在六种系统 / 架构上各自打包,并把安装包和 `SHA256SUMS` 发到 GitHub Release。签名与公证所需的仓库密钥见该文件开头的注释。同一个标签还会触发 `.github/workflows/docker.yml`,把 x64 + arm64 镜像发布到 `ghcr.io/modelflare-ai/anymodel-bot`(`<版本>` 和 `latest` 两个标签)。
## Docker
+用户直接用已发布的镜像:`docker run -d --name amb -p 3600:3600 -v amb-data:/data ghcr.io/modelflare-ai/anymodel-bot`。下面是从源码自己构建:
+
```bash
git submodule update --init --depth 1 vendor/magpie
docker build -t anymodel-bot .
diff --git a/docs/development_en.md b/docs/development_en.md
index 25aeeba..8b2fd01 100644
--- a/docs/development_en.md
+++ b/docs/development_en.md
@@ -66,10 +66,12 @@ npm run test:dsh-bundle # real DSH integration test under Electron against the
- The console listens on 127.0.0.1 only and generates a random access token on every launch, so other local programs cannot call the management API directly.
- On quit, channels stop first, then magpie. If a previous run was killed and left a magpie process behind, the next launch cleans it up (Linux/macOS).
-Releases: bump `version` in `package.json`, commit, and push a `v<version>` tag. `.github/workflows/release.yml` builds on each of the six OS / architecture runners and publishes the installers plus `SHA256SUMS` to a GitHub Release. The repository secrets needed for signing and notarization are described in the comment at the top of that file.
+Releases: bump `version` in `package.json`, commit, and push a `v<version>` tag. `.github/workflows/release.yml` builds on each of the six OS / architecture runners and publishes the installers plus `SHA256SUMS` to a GitHub Release. The repository secrets needed for signing and notarization are described in the comment at the top of that file. The same tag also starts `.github/workflows/docker.yml`, which publishes the x64 + arm64 image to `ghcr.io/modelflare-ai/anymodel-bot` (tagged `<version>` and `latest`).
## Docker
+Users run the published image: `docker run -d --name amb -p 3600:3600 -v amb-data:/data ghcr.io/modelflare-ai/anymodel-bot`. To build it from source yourself:
+
```bash
git submodule update --init --depth 1 vendor/magpie
docker build -t anymodel-bot .