Buildlogai MCP

MCPAI代理

提供方 buildlogai

AI代理 未标注 会话登录 社区 / 第三方 Claude DesktopClaudeCursorVS Code

MCP服务器为AI代理提供搜索、跟随和记录编程工作流的功能,支持自动化记录和工作流分享。

一句话选型结论

按资料字段整理

Buildlogai MCP 属于AI代理类 MCP 服务器,传输方式未标注,会话登录,可直接接入 Claude Desktop、Claude、Cursor 等,没有现成安装命令,需按官方文档接入。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类AI代理 传输方式未标注 鉴权会话登录 部署方式未标注
GitHub 星标 0 仓库热度,无公开仓库时显示未标注
工具数 11 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

AI代理
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量11暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

编程辅助TypeScript本地部署AI代理工作流管理自动化记录知识共享

安装与配置

未标注 会话登录
上游未给出传输方式
  1. 1
    准备运行环境

    按官方文档确认运行环境(常见为 Node.js 或 Python),装好后再执行安装命令。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后客户端会列出约 11 个工具,随便调一个只读工具确认能返回结果。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
给代理接外部能力

让代理在任务中途调用搜索、数据库等真实工具。

多步任务编排

把一串操作拆成步骤,由代理按顺序执行。

自动化重复流程

把固定流程交给代理跑,减少手工操作。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 资料标注 11 个工具,覆盖面较广,能满足多种调用。
  • 已适配 Claude Desktop、Claude、Cursor,换客户端不用重新研究配置。

先想清楚的地方

  • 没有现成安装命令,需要按官方文档自己填启动命令和参数。
  • 使用前要准备好会话登录对应的凭据,并按最小权限配置。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在AI代理场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
Buildlogai MCP当前 11 未标注 未标注 无 2026-05-17
Filesystem 560c4363 13 未标注 本地部署 无 2026-05-17
Server Everything (everything) 11 stdio · 本地进程 本地部署 有 2026-05-17
Time 7365cb6d 2 SSE · 服务端推送 远程部署 有 2026-05-18
Everything Claude Code 未标注 stdio · 本地进程 本地部署 有 2026-05-17
Openmemory 557e1585 未标注 stdio · 本地进程 本地部署 有 2026-05-17

服务器档案

AI代理
字段来自项目仓库与公开索引,中文说明由本站标注
提供方buildlogai上游标注的维护方 buildlogai
分类AI代理本站按上游分类展示
连接方式未标注上游未给出传输方式
鉴权方式会话登录远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · 未标注本地安装时用到的包标识与运行时
部署类型未标注上游给出的部署形态
最近核验2026-05-17上游最后核验时间

详细介绍

6.9 KB
来自项目仓库与上游目录的原文,未做删改

@buildlogai/mcp

MCP server for Buildlog - search, follow, and record AI coding workflows.

This server gives AI agents direct access to the buildlog ecosystem, enabling them to:
- Search for relevant workflows before starting a task
- Follow proven buildlog recipes in their current context
- Record their own sessions as they work (auto-record by default!)
- Upload completed buildlogs for others to learn from

Installation```bash

npm install -g @buildlogai/mcp


## Configuration

Set your API key for uploads (optional for search/follow, required for upload):```bash
export BUILDLOG_API_KEY=your-api-key

Auto-Recording

By default, the MCP server starts recording automatically when loaded. This ensures no workflow is lost.

To disable auto-recording:```bash
export BUILDLOG_AUTO_RECORD=false


When a session ends with meaningful content (3+ steps), the server will suggest uploading.

Use `buildlog_auto_status` to check if auto-recording is active.

## Proactive Suggestions

Before starting a complex task, use `buildlog_suggest`:```
Use buildlog_suggest with taskDescription "Add Stripe subscription checkout to Next.js app"

The server will search for relevant workflows and recommend the best matches.

Source Attribution

All buildlogs created via MCP include source metadata:
- Tool name and version (@buildlogai/mcp)
- Client application (if detectable via MCP_CLIENT env var)

This helps track which tools contribute to the collective knowledge base.

Usage

Claude Desktop

Add to your Claude Desktop config (~/.claude/claude_desktop_config.json):```json
{
"mcpServers": {
"buildlog": {
"command": "buildlog-mcp"
}
}
}


### Cursor

Add to your Cursor MCP settings:```json
{
  "mcpServers": {
    "buildlog": {
      "command": "npx",
      "args": ["@buildlogai/mcp"]
    }
  }
}

OpenClaw

OpenClaw supports MCP servers natively:```json
{
"mcp": {
"servers": ["buildlog-mcp"]
}
}


## Available Tools

| Tool | Description |
|------|-------------|
| `buildlog_search` | Search buildlog.ai for relevant workflows |
| `buildlog_suggest` | Get proactive workflow suggestions before starting a task |
| `buildlog_get` | Fetch a specific buildlog by slug or URL |
| `buildlog_get_steps` | Get just the steps from a buildlog |
| `buildlog_follow` | Get prompts formatted for execution |
| `buildlog_record_start` | Begin recording a session (or convert auto-session) |
| `buildlog_record_step` | Log a step (prompt, action, terminal, note) |
| `buildlog_record_stop` | End recording, get the buildlog |
| `buildlog_auto_status` | Check auto-recording status and stats |
| `buildlog_upload` | Upload to buildlog.ai |
| `buildlog_fork` | Start recording from an existing buildlog |

## Tool Details

### buildlog_search

Search buildlog.ai for relevant workflows.```
Parameters:
- query (required): Search query (e.g., "stripe nextjs integration")
- language (optional): Filter by language (e.g., "typescript")
- framework (optional): Filter by framework (e.g., "nextjs")
- limit (optional): Max results (default: 10)

buildlog_suggest

Get proactive workflow suggestions before starting a task. Call this before starting complex work.```
Parameters:
- taskDescription (required): Description of what you're about to do
- contextFiles (optional): Files relevant to the task (helps improve matching)

Returns:
- Ranked list of relevant workflows
- Top recommendation with follow command
- Extracted keywords that were matched


### buildlog_get

Fetch a complete buildlog by slug or URL.```
Parameters:
- slug (required): Buildlog slug or full URL

buildlog_get_steps

Get just the steps from a buildlog (lighter response).```
Parameters:
- slug (required): Buildlog slug or full URL


### buildlog_follow

Get prompts from a buildlog formatted for execution. This is the key tool for replicating workflows.```
Parameters:
- slug (required): Buildlog slug or full URL
- step (optional): Start from a specific step number

buildlog_record_start

Begin recording a new session. If auto-recording is active, this converts the auto-session to a named session.```
Parameters:
- title (required): Title for the recording
- description (optional): Longer description


### buildlog_record_step

Log a step to the active recording.```
Parameters:
- type (required): "prompt" | "action" | "terminal" | "note"
- content (required): The step content
- metadata (optional): Additional data (filesCreated, filesModified, etc.)

buildlog_record_stop

End recording and return the buildlog.```
Parameters:
- outcome (optional): "success" | "partial" | "failure"
- summary (optional): Summary of what was accomplished


### buildlog_auto_status

Check if auto-recording is active and get current session stats.```
Parameters: none

Returns:
- autoRecordEnabled: Whether auto-record is on
- isRecording: Whether currently recording
- isAutoSession: Whether this is an auto-started session
- stats: Step count, duration, prompt/action counts

buildlog_upload

Upload a buildlog to buildlog.ai.```
Parameters:
- buildlog (required): The buildlog object
- public (optional): Whether publicly visible (default: true)


### buildlog_fork

Start a new recording based on an existing buildlog.```
Parameters:
- slug (required): Source buildlog slug
- title (required): Title for the new buildlog
- fromStep (optional): Only copy steps up to this number

Example Workflows

Search and Follow```

Agent: Use buildlog_search to find "stripe nextjs integration"
→ Returns list of relevant buildlogs

Agent: Use buildlog_follow on buildlog "abc123"
→ Returns prompts to execute in order

Agent: Execute each prompt, adapting to current project


### Record a Session```
Agent: Use buildlog_record_start with title "Building auth system"

Agent: ... does work ...
Agent: Use buildlog_record_step for each significant action

Agent: Use buildlog_record_stop when done
→ Returns completed buildlog

Agent: Use buildlog_upload to share
→ Returns URL to view buildlog

Fork and Extend```

Agent: Use buildlog_fork on "abc123" with title "Extended auth with MFA"
→ Starts recording with inherited steps

Agent: Use buildlog_record_step for new work
Agent: Use buildlog_record_stop
Agent: Use buildlog_upload
```

Environment Variables

Variable Description
BUILDLOG_API_KEY API key for uploads (get from buildlog.ai/settings)
BUILDLOG_API_URL Custom API URL (default: https://buildlog.ai/api)
BUILDLOG_AUTO_RECORD Set to "false" to disable auto-recording (default: true)
MCP_CLIENT Client name for source attribution (e.g., "Claude Desktop", "Cursor")

VS Code Extension Interop

When recording, steps are also written to ~/.buildlog/agent-feed.jsonl (if the directory exists). This allows the VS Code extension to incorporate MCP-based recordings into its capture flow.

Links

  • buildlog.ai
  • Documentation
  • GitHub
  • VS Code Extension

License

MIT

常见问题

6 问
答案由本站依据条目字段整理,不引用上游原文
Buildlogai MCP 怎么安装?

资料没有给出安装命令,请到官方仓库或文档按说明接入;这类条目多数是生态应用,或尚未发布可安装包。

Buildlogai MCP 怎么在 Claude Desktop、Claude、Cursor 里配置?

它的传输方式未标注,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

Buildlogai MCP 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

Buildlogai MCP 需要 API Key 吗?

资料标注需要会话登录认证,使用前准备好对应凭据,并按最小权限配置。

Buildlogai MCP 支持哪些客户端?

资料标注支持 Claude Desktop、Claude、Cursor、VS Code;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

Buildlogai MCP 能提供哪些能力?

资料显示约 11 个工具;能力标签包括 编程辅助、工作流管理、TypeScript、Claude。具体工具清单以官方文档为准。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-05-17。