1 - 介绍
autogpt 介绍
AutoGPT 简介(AI)
AutoGPT 是一个开源的自主 AI Agent 框架，主要用于执行复杂的多步骤任务和自动化工作流。
起源：由开发者 Toran Bruce Richards 在 2023 年发布，最初基于 OpenAI 的 GPT‑4 模型。
核心理念：让大语言模型（LLM）不仅回答问题，还能 自主决定下一步行动，并把执行结果反馈给自己，形成迭代循环。
功能定位：
-
自动化复杂任务：用户只需给出一个高层目标，AutoGPT 会拆解成子任务并逐步完成。
-
多步骤工作流：比如市场调研、代码编写、数据分析，它能自动调用工具和 API。
-
自主性：相比 ChatGPT 需要人类不断输入提示，AutoGPT 可以自己规划和执行。
应用场景：
-
商业分析：自动收集信息、生成报告。
-
软件开发：编写和测试代码。
-
研究任务：长时间运行的知识探索。
-
个人助理：自动化日常任务，如邮件处理、日程安排。
2.1 - 概述
autogpt 概述
AutoGPT 平台是一个突破性的系统，它革新了企业和个人对 AI 的利用方式。它能够创建、部署和管理持续工作的智能代理，为您的日常工作流程带来前所未有的效率和创新能力。
主要功能
-
无缝集成和低代码工作流：无需丰富的编码知识即可快速创建复杂工作流。
-
自主运行和持续代理：部署可无限运行、在相关触发条件下激活的云端助手。
-
智能自动化和最高效率：通过自动化重复流程来简化工作流。
-
可靠性能和可预测执行：享受持续且可靠的长时间运行流程。
平台架构
AutoGPT 平台由两个主要组件组成：
-
AutoGPT 服务器
我们平台的强大核心，包含：
-
源代码：驱动代理和自动化流程的核心逻辑。
-
基础设施：确保可靠和可扩展性能的强大系统。
-
市场(Marketplace:)：一个用于预构建代理的全面市场。
-
-
AutoGPT 前端
与平台交互的用户界面：
-
Agent 构建器：设计和配置您自己的 AI 智能体。
-
工作流管理：构建、修改和优化自动化工作流。
-
部署控制：管理您的代理生命周期。
-
即用型代理：从预配置的代理中选择。
-
代理交互：通过用户友好的界面运行和交互代理。
-
监控和分析：跟踪代理性能并获取洞察。
-
平台组件
Agents 和 Workflows
在平台中，您可以创建高度定制的工作流来构建代理。代理本质上是一个您设计用于执行特定任务或流程的自动化工作流。创建定制工作流来构建用于各种任务的代理，包括：
-
数据处理和分析
-
任务调度和管理
-
通信和通知系统
-
不同软件工具之间的集成
-
AI 驱动的决策和内容生成
Blocks as Integrations
模块代表操作，是您工作流的构建模块，包括：
-
与外部服务的连接
-
数据处理工具
-
用于各种任务的 AI 模型
-
自定义脚本或函数
-
条件逻辑和决策组件
可用语言模型
该平台预集成了前沿的 LLM 提供商：
- OpenAI - https://openai.com/
- Anthropic - https://www.anthropic.com/
- Groq - https://groq.com/
- Llama - https://llamaindex.ai/
- AI/ML API - https://aimlapi.com/
- AI/ML API 提供 300 多个 AI 模型，包括 Deepseek、Gemini 和 ChatGPT。这些模型运行在企业级速率限制和正常运行时间标准下。
3 - 启动速度
autogpt Agent 的启动速度
部署
参考:
https://docs.agpt.co/platform/getting-started/
准备工作
这三个我 linux mint 机器上都有:
- Node.js
- Docker
- Git
安装
使用自动安装脚本进行安装:
mkdir -p ~/work/code/agents
cd ~/work/code/agents
curl -fsSL https://setup.agpt.co/install.sh -o install.sh && bash install.sh
输出为:
d8888 888 .d8888b. 8888888b. 88888888888
d88888 888 d88P Y88b 888 Y88b 888
d88P888 888 888 888 888 888 888
d88P 888 888 888 888888 .d88b. 888 888 d88P 888
d88P 888 888 888 888 d88""88b 888 88888 8888888P" 888
d88P 888 888 888 888 888 888 888 888 888 888
d8888888888 Y88b 888 Y88b. Y88..88P Y88b d88P 888 888
d88P 888 "Y88888 "Y888 "Y88P" "Y8888P88 888 888
AutoGPT Setup Script
-------------------
Checking prerequisites...
✓ Git is installed
✓ Docker is installed
All prerequisites installed!
Cloning AutoGPT repository...
Cloning into '/home/sky/work/code/agents/AutoGPT'...
remote: Enumerating objects: 129027, done.
remote: Counting objects: 100% (604/604), done.
remote: Compressing objects: 100% (359/359), done.
remote: Total 129027 (delta 471), reused 249 (delta 245), pack-reused 128423 (from 4)
Receiving objects: 100% (129027/129027), 297.76 MiB | 17.41 MiB/s, done.
Resolving deltas: 100% (83388/83388), done.
Repository cloned successfully.
Starting AutoGPT services with Docker Compose...
This may take a few minutes on first run...
✓ Services started successfully!
=============================
Setup Complete!
=============================
🚀 Access AutoGPT at: http://localhost:3000
📡 API available at: http://localhost:8000
To stop services: docker compose down
To view logs: docker compose logs -f
All commands should be run in: /home/sky/work/code/agents/AutoGPT/autogpt_platform
安装过程中自动 clone AutoGPT 代码仓库.
配置
cd ~/work/code/agents/AutoGPT/autogpt_platform
cp .env.default .env
启动
第一次:
cd ~/work/code/agents/AutoGPT
docker compose up -d --build
输出为:
docker compose up -d --build
Compose can now delegate builds to bake for better performance.
To do so, set COMPOSE_BAKE=true.
[+] Building 18.8s (132/220) docker:default
=> [migrate internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [notification_server internal] load metadata for docker.io/library/debian:13-slim 18.1s
=> [migrate internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [migrate internal] load build context 0.0s
=> => transferring context: 1.03MB 0.0s
=> [notification_server builder 1/13] FROM docker.io/library/debian:13-slim@sha256:18764 0.0s
=> CACHED [websocket_server builder 2/13] WORKDIR /app 0.0s
=> CACHED [websocket_server server_dependencies 3/16] RUN apt-get update && apt-get inst 0.0s
=> CACHED [websocket_server builder 3/13] RUN echo 'Acquire::http::Pipeline-Depth 0;\nAc 0.0s
=> CACHED [websocket_server builder 4/13] RUN apt-get update --allow-releaseinfo-change 0.0s
=> CACHED [websocket_server builder 5/13] RUN apt-get update && apt-get install -y 0.0s
=> CACHED [websocket_server builder 6/13] RUN pip3 install poetry --break-system-package 0.0s
=> CACHED [migrate builder 7/13] COPY autogpt_platform/autogpt_libs /app/autogpt_platfor 0.0s
=> CACHED [migrate builder 8/13] COPY autogpt_platform/backend/poetry.lock autogpt_platf 0.0s
=> CACHED [migrate builder 9/13] WORKDIR /app/autogpt_platform/backend 0.0s
=> CACHED [migrate builder 10/13] RUN poetry install --no-ansi --no-root 0.0s
=> CACHED [migrate builder 11/13] COPY autogpt_platform/backend/schema.prisma ./ 0.0s
=> CACHED [migrate builder 12/13] COPY autogpt_platform/backend/backend/data/partial_type 0.0s
=> CACHED [migrate builder 13/13] RUN poetry run prisma generate 0.0s
=> CACHED [migrate server_dependencies 4/16] COPY --from=builder /app /app 0.0s
=> CACHED [migrate server_dependencies 5/16] COPY --from=builder /usr/local/lib/python3* 0.0s
=> CACHED [migrate server_dependencies 6/16] COPY --from=builder /usr/local/bin/poetry / 0.0s
=> CACHED [migrate server_dependencies 7/16] COPY --from=builder /usr/bin/node /usr/bin/ 0.0s
=> CACHED [migrate server_dependencies 8/16] COPY --from=builder /usr/lib/node_modules / 0.0s
=> CACHED [migrate server_dependencies 9/16] COPY --from=builder /usr/bin/npm /usr/bin/n 0.0s
=> CACHED [migrate server_dependencies 10/16] COPY --from=builder /usr/bin/npx /usr/bin/n 0.0s
=> CACHED [migrate server_dependencies 11/16] COPY --from=builder /root/.cache/prisma-pyt 0.0s
=> CACHED [migrate server_dependencies 12/16] RUN mkdir -p /app/autogpt_platform/autogpt_ 0.0s
=> CACHED [migrate server_dependencies 13/16] RUN mkdir -p /app/autogpt_platform/backend 0.0s
=> CACHED [migrate server_dependencies 14/16] COPY autogpt_platform/autogpt_libs /app/aut 0.0s
=> CACHED [migrate server_dependencies 15/16] COPY autogpt_platform/backend/poetry.lock a 0.0s
=> CACHED [migrate server_dependencies 16/16] WORKDIR /app/autogpt_platform/backend 0.0s
=> CACHED [migrate migrate 1/3] COPY autogpt_platform/backend/schema.prisma /app/autogpt_ 0.0s
=> CACHED [migrate migrate 2/3] COPY autogpt_platform/backend/backend/data/partial_types. 0.0s
=> CACHED [migrate migrate 3/3] COPY autogpt_platform/backend/migrations /app/autogpt_pla 0.0s
=> [migrate] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:79e4f5dc5d9d3fd8d10aa986e3a367ff3525af72c047cd53dc7c740db21989 0.0s
=> => naming to docker.io/library/autogpt_platform-migrate 0.0s
=> [migrate] resolving provenance for metadata file 0.0s
=> [database_manager internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [frontend internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 1.96kB 0.0s
=> [rest_server internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [frontend internal] load metadata for docker.io/library/node:21-alpine 6.6s
=> [rest_server internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [database_manager internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [database_manager internal] load build context 0.1s
=> => transferring context: 4.64MB 0.1s
=> [rest_server internal] load build context 0.1s
=> => transferring context: 5.64MB 0.1s
=> CACHED [database_manager builder 7/13] COPY autogpt_platform/autogpt_libs /app/autogp 0.0s
=> CACHED [database_manager builder 8/13] COPY autogpt_platform/backend/poetry.lock auto 0.0s
=> CACHED [database_manager builder 9/13] WORKDIR /app/autogpt_platform/backend 0.0s
=> CACHED [database_manager builder 10/13] RUN poetry install --no-ansi --no-root 0.0s
=> CACHED [database_manager builder 11/13] COPY autogpt_platform/backend/schema.prisma ./ 0.0s
=> CACHED [database_manager builder 12/13] COPY autogpt_platform/backend/backend/data/par 0.0s
=> CACHED [database_manager builder 13/13] RUN poetry run prisma generate 0.0s
=> CACHED [database_manager server_dependencies 4/16] COPY --from=builder /app /app 0.0s
=> CACHED [database_manager server_dependencies 5/16] COPY --from=builder /usr/local/lib 0.0s
=> CACHED [database_manager server_dependencies 6/16] COPY --from=builder /usr/local/bin 0.0s
=> CACHED [database_manager server_dependencies 7/16] COPY --from=builder /usr/bin/node 0.0s
=> CACHED [database_manager server_dependencies 8/16] COPY --from=builder /usr/lib/node_ 0.0s
=> CACHED [database_manager server_dependencies 9/16] COPY --from=builder /usr/bin/npm / 0.0s
=> CACHED [database_manager server_dependencies 10/16] COPY --from=builder /usr/bin/npx / 0.0s
=> CACHED [database_manager server_dependencies 11/16] COPY --from=builder /root/.cache/p 0.0s
=> CACHED [database_manager server_dependencies 12/16] RUN mkdir -p /app/autogpt_platform 0.0s
=> CACHED [database_manager server_dependencies 13/16] RUN mkdir -p /app/autogpt_platform 0.0s
=> CACHED [database_manager server_dependencies 14/16] COPY autogpt_platform/autogpt_libs 0.0s
=> CACHED [database_manager server_dependencies 15/16] COPY autogpt_platform/backend/poet 0.0s
=> CACHED [database_manager server_dependencies 16/16] WORKDIR /app/autogpt_platform/back 0.0s
=> CACHED [database_manager server 1/2] COPY autogpt_platform/backend /app/autogpt_platfo 0.0s
=> CACHED [rest_server server 2/2] RUN poetry install --no-ansi --only-root 0.0s
=> [database_manager] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:793d80d2bcf86a4d20e45d8378731ccfe50afa0b3555b1379a86b617e51962 0.0s
=> => naming to docker.io/library/autogpt_platform-database_manager 0.0s
=> [rest_server] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:29521f5b6ab2cf880eb2f39bb4957f4ab3a5a93bd09ed2f6ee5d87ab1703b0 0.0s
=> => naming to docker.io/library/autogpt_platform-rest_server 0.0s
=> [database_manager] resolving provenance for metadata file 0.0s
=> [rest_server] resolving provenance for metadata file 0.0s
=> [notification_server internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [websocket_server internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [executor internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [scheduler_server internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 3.80kB 0.0s
=> [frontend internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [frontend base 1/5] FROM docker.io/library/node:21-alpine@sha256:78c45726ea205bbe2f238 0.0s
=> [frontend internal] load build context 0.2s
=> => transferring context: 20.79MB 0.2s
=> CACHED [frontend base 2/5] WORKDIR /app 0.0s
=> CACHED [frontend prod 3/9] RUN addgroup --system --gid 1001 nodejs 0.0s
=> CACHED [frontend prod 4/9] RUN adduser --system --uid 1001 nextjs 0.0s
=> CACHED [frontend prod 5/9] RUN mkdir .next 0.0s
=> CACHED [frontend prod 6/9] RUN chown nextjs:nodejs .next 0.0s
=> CACHED [frontend base 3/5] RUN corepack enable 0.0s
=> CACHED [frontend base 4/5] COPY autogpt_platform/frontend/package.json autogpt_platfor 0.0s
=> CACHED [frontend base 5/5] RUN --mount=type=cache,target=/root/.local/share/pnpm pnpm 0.0s
=> CACHED [frontend build 1/4] COPY autogpt_platform/frontend/ . 0.0s
=> CACHED [frontend build 2/4] RUN if [ -f .env.production ]; then cat .env.default 0.0s
=> CACHED [frontend build 3/4] RUN pnpm run generate:api 0.0s
=> CACHED [frontend build 4/4] RUN if [ "false" = "true" ]; then NEXT_PUBLIC_PW_TEST=true 0.0s
=> CACHED [frontend prod 7/9] COPY --from=build --chown=nextjs:nodejs /app/.next/standalo 0.0s
=> CACHED [frontend prod 8/9] COPY --from=build --chown=nextjs:nodejs /app/.next/static . 0.0s
=> CACHED [frontend prod 9/9] COPY --from=build /app/public ./public 0.0s
=> [frontend] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:1c123401968788aa915e2ee9ed20010c36cdf917da95c7c57f0baa7736cc43 0.0s
=> => naming to docker.io/library/autogpt_platform-frontend 0.0s
=> [frontend] resolving provenance for metadata file 0.0s
=> [scheduler_server internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [notification_server internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [websocket_server internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [executor internal] load .dockerignore 0.0s
=> => transferring context: 1.85kB 0.0s
=> [scheduler_server internal] load build context 0.1s
=> => transferring context: 5.64MB 0.1s
=> [websocket_server internal] load build context 0.0s
=> => transferring context: 65.49kB 0.0s
=> [executor internal] load build context 0.1s
=> => transferring context: 5.64MB 0.1s
=> [notification_server internal] load build context 0.1s
=> => transferring context: 5.64MB 0.1s
=> CACHED [websocket_server builder 7/13] COPY autogpt_platform/autogpt_libs /app/autogp 0.0s
=> CACHED [websocket_server builder 8/13] COPY autogpt_platform/backend/poetry.lock auto 0.0s
=> CACHED [websocket_server builder 9/13] WORKDIR /app/autogpt_platform/backend 0.0s
=> CACHED [websocket_server builder 10/13] RUN poetry install --no-ansi --no-root 0.0s
=> CACHED [websocket_server builder 11/13] COPY autogpt_platform/backend/schema.prisma ./ 0.0s
=> CACHED [websocket_server builder 12/13] COPY autogpt_platform/backend/backend/data/par 0.0s
=> CACHED [websocket_server builder 13/13] RUN poetry run prisma generate 0.0s
=> CACHED [websocket_server server_dependencies 4/16] COPY --from=builder /app /app 0.0s
=> CACHED [websocket_server server_dependencies 5/16] COPY --from=builder /usr/local/lib 0.0s
=> CACHED [websocket_server server_dependencies 6/16] COPY --from=builder /usr/local/bin 0.0s
=> CACHED [websocket_server server_dependencies 7/16] COPY --from=builder /usr/bin/node 0.0s
=> CACHED [websocket_server server_dependencies 8/16] COPY --from=builder /usr/lib/node_ 0.0s
=> CACHED [websocket_server server_dependencies 9/16] COPY --from=builder /usr/bin/npm / 0.0s
=> CACHED [websocket_server server_dependencies 10/16] COPY --from=builder /usr/bin/npx / 0.0s
=> CACHED [websocket_server server_dependencies 11/16] COPY --from=builder /root/.cache/p 0.0s
=> CACHED [websocket_server server_dependencies 12/16] RUN mkdir -p /app/autogpt_platform 0.0s
=> CACHED [websocket_server server_dependencies 13/16] RUN mkdir -p /app/autogpt_platform 0.0s
=> CACHED [websocket_server server_dependencies 14/16] COPY autogpt_platform/autogpt_libs 0.0s
=> CACHED [websocket_server server_dependencies 15/16] COPY autogpt_platform/backend/poet 0.0s
=> CACHED [websocket_server server_dependencies 16/16] WORKDIR /app/autogpt_platform/back 0.0s
=> CACHED [websocket_server server 1/2] COPY autogpt_platform/backend /app/autogpt_platfo 0.0s
=> CACHED [scheduler_server server 2/2] RUN poetry install --no-ansi --only-root 0.0s
=> [websocket_server] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:2704efb2a355df26e3eeac826ec53efe576464affd72ba10cc4fd58993949c 0.0s
=> => naming to docker.io/library/autogpt_platform-websocket_server 0.0s
=> [websocket_server] resolving provenance for metadata file 0.0s
=> [executor] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:e1c89cc61ecf7c1e1e0277cb9821019977ba3332b530015372e8c218944fa7 0.0s
=> => naming to docker.io/library/autogpt_platform-executor 0.0s
=> [scheduler_server] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:c52a47b1f606c5c9cddc813d480b3953fa7c98ba308d453edea728eaefa6e8 0.0s
=> => naming to docker.io/library/autogpt_platform-scheduler_server 0.0s
=> [notification_server] exporting to image 0.0s
=> => exporting layers 0.0s
=> => writing image sha256:12b086866f50c684190215410e0cf4587e2d95c9f95acbd0121592ed4bea66 0.0s
=> => naming to docker.io/library/autogpt_platform-notification_server 0.0s
=> [executor] resolving provenance for metadata file 0.0s
=> [scheduler_server] resolving provenance for metadata file 0.0s
=> [notification_server] resolving provenance for metadata file 0.0s
[+] Running 22/22
✔ database_manager Built 0.0s
✔ executor Built 0.0s
✔ frontend Built 0.0s
✔ migrate Built 0.0s
✔ notification_server Built 0.0s
✔ rest_server Built 0.0s
✔ scheduler_server Built 0.0s
✔ websocket_server Built 0.0s
✔ Container supabase-kong Running 0.0s
✔ Container rabbitmq Healthy 4.3s
✔ Container supabase-db Healthy 4.3s
✔ Container autogpt_platform-redis-1 Healthy 4.3s
✔ Container autogpt_platform-clamav-1 Running 0.0s
✔ Container autogpt_platform-migrate-1 Exited 4.3s
✔ Container supabase-auth Running 0.0s
✔ Container autogpt_platform-rest_server-1 Running 0.0s
✔ Container autogpt_platform-frontend-1 Running 0.0s
✔ Container autogpt_platform-database_manager-1 Running 0.0s
✔ Container autogpt_platform-websocket_server-1 Running 0.0s
✔ Container autogpt_platform-scheduler_server-1 Running 0.0s
✔ Container autogpt_platform-executor-1 Running 0.0s
✔ Container autogpt_platform-notification_server-1 Running
此时执行:
docker ps
可以看到 autogpt 的多个容器和依赖的底层组件:
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
cb7a557cf896 autogpt_platform-notification_server "python -m backend.n…" 3 minutes ago Up 3 minutes 0.0.0.0:8007->8007/tcp, [::]:8007->8007/tcp autogpt_platform-notification_server-1
3b55b78d16e8 autogpt_platform-websocket_server "python -m backend.ws" 3 minutes ago Up 3 minutes 0.0.0.0:8001->8001/tcp, [::]:8001->8001/tcp autogpt_platform-websocket_server-1
300ae427cb8f autogpt_platform-scheduler_server "python -m backend.s…" 3 minutes ago Up 3 minutes 0.0.0.0:8003->8003/tcp, [::]:8003->8003/tcp autogpt_platform-scheduler_server-1
142479a30cf1 autogpt_platform-executor "python -m backend.e…" 3 minutes ago Up 3 minutes 0.0.0.0:8002->8002/tcp, [::]:8002->8002/tcp autogpt_platform-executor-1
96cb2f3ebf36 autogpt_platform-rest_server "python -m backend.r…" 3 minutes ago Up 3 minutes 0.0.0.0:8006->8006/tcp, [::]:8006->8006/tcp autogpt_platform-rest_server-1
39234cb8a67f autogpt_platform-database_manager "python -m backend.db" 3 minutes ago Up 3 minutes 0.0.0.0:8005->8005/tcp, [::]:8005->8005/tcp autogpt_platform-database_manager-1
aa403ad8f61f supabase/postgres:15.8.1.049 "docker-entrypoint.s…" 3 minutes ago Up 3 minutes (healthy) 0.0.0.0:5432->5432/tcp, [::]:5432->5432/tcp supabase-db
2681b4df8b73 kong:2.8.1 "bash -c 'eval \"echo…" 3 minutes ago Up 3 minutes (healthy) 0.0.0.0:8000->8000/tcp, [::]:8000->8000/tcp, 8001/tcp, 0.0.0.0:8443->8443/tcp, [::]:8443->8443/tcp, 8444/tcp supabase-kong
c2f2d97c0cde autogpt_platform-frontend "docker-entrypoint.s…" 28 minutes ago Up 28 minutes 0.0.0.0:3000->3000/tcp, [::]:3000->3000/tcp autogpt_platform-frontend-1
7ca46aa8769f supabase/gotrue:v2.170.0 "auth" 28 minutes ago Up 28 minutes (healthy) supabase-auth
cad37656bf28 rabbitmq:management "docker-entrypoint.s…" 28 minutes ago Up 28 minutes (healthy) 4369/tcp, 5671/tcp, 0.0.0.0:5672->5672/tcp, [::]:5672->5672/tcp, 15671/tcp, 15691-15692/tcp, 25672/tcp, 0.0.0.0:15672->15672/tcp, [::]:15672->15672/tcp rabbitmq
454f81fbb681 redis:latest "docker-entrypoint.s…" 28 minutes ago Up 28 minutes (healthy) 0.0.0.0:6379->6379/tcp, [::]:6379->6379/tcp autogpt_platform-redis-1
3e0e32178353 clamav/clamav-debian:latest "/init" 28 minutes ago Up 28 minutes (healthy) 0.0.0.0:3310->3310/tcp, [::]:3310->3310/tcp, 7357/tcp autogpt_platform-clamav-1
此时访问 http://localhost:3000/ 就可以看到 autogpt 的页面.
以后再启动, 不需要再 build 的
docker compose up -d
启动速度
为了测试启动时间,前后打印一下时间值:
TZ=UTC-8 date +"%Y-%m-%d %H:%M:%S,%3N"
docker compose up -d
TZ=UTC-8 date +"%Y-%m-%d %H:%M:%S,%3N"
输出为:
2025-11-20 18:02:46,972
[+] Running 16/16
✔ Network shared-network Created 0.0s
✔ Network app-network Created 0.0s
✔ Container rabbitmq Healthy 10.9s
✔ Container supabase-db Healthy 10.9s
✔ Container autogpt_platform-clamav-1 Started 0.4s
✔ Container supabase-kong Started 0.4s
✔ Container autogpt_platform-redis-1 Healthy 10.6s
✔ Container autogpt_platform-migrate-1 Exited 10.8s
✔ Container supabase-auth Started 6.1s
✔ Container autogpt_platform-frontend-1 Started 10.3s
✔ Container autogpt_platform-database_manager-1 Started 10.3s
✔ Container autogpt_platform-rest_server-1 Started 10.7s
✔ Container autogpt_platform-websocket_server-1 Started 11.0s
✔ Container autogpt_platform-notification_server-1 Started 11.1s
✔ Container autogpt_platform-executor-1 Started 11.0s
✔ Container autogpt_platform-scheduler_server-1 Started 11.0s
2025-11-20 18:02:58,334
启动时长为 11.36 秒. 重复几次, 11.247 / 11.353, 取 11.35 秒.
内存占用
启动之后, 执行命令:
docker ps --format "{{.Names}}" | xargs -I {} docker stats {} --no-stream --format "{{.Name}}: {{.MemUsage}}"
能看到各个容器的内存占用:
autogpt_platform-notification_server-1: 212.3MiB / 31.11GiB
autogpt_platform-executor-1: 264MiB / 31.11GiB
autogpt_platform-websocket_server-1: 206.5MiB / 31.11GiB
autogpt_platform-scheduler_server-1: 270.1MiB / 31.11GiB
autogpt_platform-rest_server-1: 513.2MiB / 31.11GiB
autogpt_platform-database_manager-1: 279.3MiB / 31.11GiB
autogpt_platform-frontend-1: 130.6MiB / 31.11GiB
supabase-auth: 8.438MiB / 31.11GiB
supabase-db: 65.4MiB / 31.11GiB
rabbitmq: 123.2MiB / 31.11GiB
autogpt_platform-redis-1: 5.168MiB / 31.11GiB
supabase-kong: 1.645GiB / 31.11GiB
autogpt_platform-clamav-1: 1.299GiB / 31.11GiB
写个脚本统计一下内存总数:
docker ps --format "{{.Names}}" | \
xargs -I {} docker stats {} --no-stream --format "{{.MemUsage}}" | \
awk '{split($1,a,"MiB"); sum+=a[1]} END {print "AutoGPT 总内存占用: " sum " MiB"}'
输出为 AutoGPT 总内存占用: 2082.09 MiB . 整整2.08GB的内存使用.
关闭
cd ~/work/code/agents/AutoGPT
docker compose down
输出:
[+] Running 16/16
✔ Container autogpt_platform-rest_server-1 Removed 2.1s
✔ Container autogpt_platform-websocket_server-1 Removed 1.2s
✔ Container autogpt_platform-frontend-1 Removed 0.5s
✔ Container supabase-auth Removed 0.4s
✔ Container autogpt_platform-executor-1 Removed 1.4s
✔ Container supabase-kong Removed 0.6s
✔ Container autogpt_platform-clamav-1 Removed 10.4s
✔ Container autogpt_platform-notification_server-1 Removed 1.4s
✔ Container autogpt_platform-scheduler_server-1 Removed 1.5s
✔ Container autogpt_platform-database_manager-1 Removed 1.2s
✔ Container autogpt_platform-redis-1 Removed 0.4s
✔ Container rabbitmq Removed 1.5s
✔ Container autogpt_platform-migrate-1 Removed 0.0s
✔ Container supabase-db Removed 1.4s
✔ Network app-network Removed 0.4s
✔ Network shared-network Removed 0.2s