Skip to main content

MCP server for MyLLMWiki (Obsidian) capture across Cursor, Codex, Claude

Project description

MyLLMWiki MCP Server

Cursor / Codex / Claude Code 的统一 wiki 能力层。MCP 进程在宿主环境运行,不受 Codex 项目沙箱限制,写入 MyLLMWiki 更可靠。

架构

任意 vibecoding 工具
    │  MCP stdio
    ▼
integrations/mcp/server.py
    │  读 config.env
    ▼
MyLLMWiki/wiki/  +  index.md  +  log.md  +  inbox.jsonl
    ▼
Obsidian

与 Skill/Rule 的关系

职责
Skill / Rule 何时沉淀(触发词、质量闸门)
MCP 如何读写(路径、搜索、写页、记 log)
CAPTURE.md 写什么(SSOT 约定)

安装

cd /Users/jack/MyLLMWiki/integrations/mcp
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

# 或从仓库根:
./integrations/install.sh mcp

工具列表

Tool 用途
wiki_paths 返回可配置路径(同 myllmwiki paths
wiki_get_spec CAPTURE.md
wiki_read_page 读任意 wiki 文件
wiki_search wiki/ 内搜索
wiki_list_recent 列出近期 summaries/projects/…
wiki_write_page 写入/追加页面(仅限受管目录)
wiki_append_log 追加 log.md 条目
wiki_inbox 追加会话索引
wiki_capture_checklist 返回沉淀步骤(用户已触发时)
wiki_update_index 插入/更新 index.md 分区表格行(v0.2)
wiki_list_by_status 按 frontmatter status 列页(v0.2)
wiki_set_status 设置页 frontmatter status(v0.2)

wiki_paths()mcp_schema_version(当前 0.2)。

PyPI

uvx myllmwiki-mcp   # 或 pip install myllmwiki-mcp

Cursor 配置

~/.cursor/mcp.json 或项目 .cursor/mcp.json

{
  "mcpServers": {
    "myllmwiki": {
      "command": "/Users/jack/MyLLMWiki/integrations/mcp/.venv/bin/python",
      "args": ["/Users/jack/MyLLMWiki/integrations/mcp/server.py"],
      "env": {
        "MYLLMWIKI_ROOT": "/Users/jack/MyLLMWiki"
      }
    }
  }
}

Codex 配置

~/.codex/config.toml

[mcp_servers.myllmwiki]
type = "stdio"
command = "/Users/jack/MyLLMWiki/integrations/mcp/.venv/bin/python"
args = ["/Users/jack/MyLLMWiki/integrations/mcp/server.py"]

[mcp_servers.myllmwiki.env]
MYLLMWIKI_ROOT = "/Users/jack/MyLLMWiki"

重启 Codex 后在工具列表中应出现 myllmwiki MCP。

Claude Code 配置

~/.claude/settings.json 或项目 .mcp.json(视版本而定):

{
  "mcpServers": {
    "myllmwiki": {
      "command": "/Users/jack/MyLLMWiki/integrations/mcp/.venv/bin/python",
      "args": ["/Users/jack/MyLLMWiki/integrations/mcp/server.py"],
      "env": {
        "MYLLMWIKI_ROOT": "/Users/jack/MyLLMWiki"
      }
    }
  }
}

推荐使用方式

用户说「沉淀」后,agent 应:

  1. wiki_capture_checklist(topics=[...], project="...")
  2. wiki_get_spec() + wiki_search(...) 避免重复建页
  3. wiki_write_page(...) 写 summary / concept / project
  4. wiki_append_log(...) + 更新 wiki/index.md

MCP 不能替代触发闸门 — 仍需要用户说「沉淀」或 rule/skill 约束,避免自动写 wiki。

可配置目录

与 CLI 相同:编辑 integrations/config.env,MCP 启动时自动加载。

本地测试

source integrations/mcp/.venv/bin/activate
python integrations/mcp/server.py   # stdio,需 MCP 客户端连接

# 或快速测 paths 模块:
python -c "from paths import load_paths; print(load_paths())"

优劣

MCP Skill + 直接写文件
跨工具 ✅ 同一 server ❌ 每工具一套
Codex 沙箱 ✅ 进程外写入 ❌ 常需 escalation
路径校验 ✅ server 内统一 ❌ 靠 prompt
实现成本
需安装 venv + 各工具注册 symlink 即可

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

myllmwiki_mcp-0.2.0.tar.gz (10.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

myllmwiki_mcp-0.2.0-py3-none-any.whl (10.6 kB view details)

Uploaded Python 3

File details

Details for the file myllmwiki_mcp-0.2.0.tar.gz.

File metadata

  • Download URL: myllmwiki_mcp-0.2.0.tar.gz
  • Upload date:
  • Size: 10.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.7

File hashes

Hashes for myllmwiki_mcp-0.2.0.tar.gz
Algorithm Hash digest
SHA256 779582e07a31dde47f78f4b847d134a55ffc56b7153bfa515fa9d230000d6779
MD5 69ab8233b8438f01ffafcee035e9d942
BLAKE2b-256 34b6985c99c8500babff4c5337ec88da814a5833f465a8f36b08298976e24424

See more details on using hashes here.

File details

Details for the file myllmwiki_mcp-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: myllmwiki_mcp-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 10.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.7

File hashes

Hashes for myllmwiki_mcp-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 0a380be6da5ef4f6f90174230988ab00ad92f044d6d1d6587fa5950b2eddaf04
MD5 eed4aed5ff1b42e5bb402452ca6f1d28
BLAKE2b-256 1f99c9b866d3603f09d10bcdcfc42109dfcfa674422756adf132b55f8c20ad0c

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page