经 AI Skill Hub 精选评估,黑盒日志 获评「强烈推荐」。这款MCP工具在功能完整性、社区活跃度和易用性方面表现出色,AI 评分 8.0 分,适合有一定技术背景的用户使用。
黑盒日志 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
黑盒日志 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/maxjb-xyz/blackbox
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"----": {
"command": "npx",
"args": ["-y", "blackbox"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 黑盒日志 执行以下任务... Claude: [自动调用 黑盒日志 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"____": {
"command": "npx",
"args": ["-y", "blackbox"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
<p align="center"> <img src="docs/assets/logo.png" alt="Blackbox" width="120" /> </p>
An intelligent self-hosted forensic event timeline for homelabs and home servers.
Know what changed, when it changed, and why things broke.
<p align="center"> <img src="https://img.shields.io/badge/Go-1.26-00798A?logo=go&logoColor=white" /> <img src="https://img.shields.io/badge/React-19-149ECA?logo=react&logoColor=white" /> <img src="https://img.shields.io/badge/SQLite-portable-002040?logo=sqlite&logoColor=white" /> <img src="https://img.shields.io/badge/Docker-ready-0B5EA8?logo=docker&logoColor=white" /> <img src="https://img.shields.io/badge/license-AGPL_v3-A00000" /> </p>
---
<img width="1502" height="810" alt="image" src="https://github.com/user-attachments/assets/a00c9afb-b10d-4bb5-b66b-13d12d846656" />
---
Development setup, local run instructions, and contributor docs live at:
---
This gets a single-node setup running in minutes. For multi-node, see the deployment docs. Both images run as non-root. The server uses UID 65532 (fixed). The agent defaults to UID/GID 65532 but can be overridden withPUID/PGIDto match the owner of your watched host paths. The agent entrypoint auto-detects the GIDs of any mounted resources (Docker socket, systemd journal) at startup - no manual group configuration needed. The project is available for bothamd64andarm64architectures.
1. Create a docker-compose.yml:
services:
blackbox-server:
image: ghcr.io/maxjb-xyz/blackbox-server:latest
container_name: blackbox-server
restart: unless-stopped
ports:
- "8080:8080"
volumes:
- blackbox-data:/data
environment:
JWT_SECRET: "change-me-to-a-long-random-string"
AGENT_TOKENS: "homelab=change-me-to-a-secret-agent-token"
WEBHOOK_SECRET: "change-me-to-a-webhook-secret"
TZ: "America/New_York" # optional: set to your local timezone so container logs match your clock
networks:
- blackbox
blackbox-agent:
image: ghcr.io/maxjb-xyz/blackbox-agent:latest
container_name: blackbox-agent
restart: unless-stopped
cap_drop:
- ALL
cap_add:
- SETUID
- SETGID
security_opt:
- no-new-privileges:true
read_only: true
volumes:
- blackbox-agent-data:/data
- /var/run/docker.sock:/var/run/docker.sock:ro
- /etc:/watch/etc:ro
- /run/log/journal:/run/log/journal:ro
- /var/log/journal:/var/log/journal:ro
- /etc/machine-id:/etc/machine-id:ro
environment:
SERVER_URL: "http://blackbox-server:8080"
AGENT_TOKEN: "change-me-to-a-secret-agent-token"
NODE_NAME: "homelab"
WATCH_PATHS: "/watch/etc"
WATCH_SYSTEMD: "true"
TZ: "America/New_York" # optional: set to your local timezone so container logs match your clock
networks:
- blackbox
volumes:
blackbox-agent-data:
blackbox-data:
networks:
blackbox:
driver: bridge
2. Start it:
docker compose up -d
3. Open http://your-server:8080 and complete the setup wizard.
4. Optional:
Admin > Data Sources to manage per-node and server-wide data sources and set up capability-aware systemd, file watcher, webhook, and Docker exclusions.Admin > System to configure file-diff redaction and Ollama-based incident analysis.---
A disclaimer - generative AI is used in the development of this repository. The agenda, features, roadmap, etc. are all set by me (a human), but a large portion of the code in this project is created by generative AI. I scan this code for issues and vulnerabilities the best I know how, but I'm not an experienced programmer.
If that makes you uncomfortable, please feel free to poke around the codebase and submit issues for anything out of place. I welcome feedback and suggestions from those more experienced than me. Please send me a private message if you find a security vulnerability that may affect other users, so I can fix it before informing everyone.
---
/mcp on the main server and protects it with a server-wide bearer token that you can regenerate from Admin. Older MCP clients configured for a separate port or /sse endpoint must be updated. Available tools: list_incidents, get_incident, list_entries, search_entries, and list_nodes.高质量的开源MCP工具,功能强大
该工具使用 AGPL-3.0 协议,商用场景请仔细阅读协议条款,必要时咨询法律意见。
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
⚠️ AGPL 3.0 — 最严格的 Copyleft,网络服务端使用也需开源,SaaS 使用受限。
AI Skill Hub 点评:黑盒日志 的核心功能完整,质量优秀。对于Claude Desktop / Claude Code 用户来说,这是一个值得纳入个人工具库的选择。建议先在非生产环境试用,再逐步推广。
| 原始名称 | blackbox |
| Topics | dockerhomelablogsmcpobservability |
| GitHub | https://github.com/maxjb-xyz/blackbox |
| License | AGPL-3.0 |
| 语言 | Go |
收录时间:2026-06-03 · 更新时间:2026-06-03 · License:AGPL-3.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端