经 AI Skill Hub 精选评估,Claude代码维基 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。
Claude代码维基 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。
Claude代码维基 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/djolex999/vir
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"claude----": {
"command": "npx",
"args": ["-y", "vir"]
}
}
}
# 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 Claude代码维基 执行以下任务... Claude: [自动调用 Claude代码维基 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": {
"claude____": {
"command": "npx",
"args": ["-y", "vir"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
<p align="center"> <img src="assets/vir_whirlpool_logo.svg" width="200" height="200" alt="vir logo"> </p>
<p align="center"> An LLM Wiki for Claude Code, in your Obsidian vault. </p>
<p align="center"> <a href="https://www.npmjs.com/package/@djolex999/vir-cli"><img src="https://img.shields.io/npm/v/@djolex999/vir-cli?color=7c6af7&label=npm" alt="npm version"></a> <a href="https://www.npmjs.com/package/@djolex999/vir-cli"><img src="https://img.shields.io/npm/dw/@djolex999/vir-cli?color=4fd1a0" alt="npm downloads"></a> <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-22d3ee" alt="license"></a> <a href="#project-status"><img src="https://img.shields.io/badge/tests-274%20passing-22c55e" alt="tests"></a> <a href="#project-status"><img src="https://img.shields.io/badge/platforms-macOS%20%7C%20Linux-lightgrey" alt="platforms"></a> <a href="https://modelcontextprotocol.io"><img src="https://img.shields.io/badge/MCP-server-c084fc" alt="mcp"></a> <a href="#"><img src="https://img.shields.io/badge/local--first-yes-f59e0b" alt="local-first"></a> <a href="https://github.com/djolex999/vir"><img src="https://img.shields.io/github/stars/djolex999/vir?style=social" alt="stars"></a> </p>
Vir is actively developed. In the next 30–60 days:
- Obsidian plugin v1 — sidebar, command palette, canvas integration, submitted to the community plugin marketplace - Multi-agent support — Codex CLI, Cursor, Aider, Cline (one per release) - PDF/paper ingestion — broaden beyond developer workflows
Track progress via GitHub issues or watch the repo for releases.
~/.claude/projects/)nomic-embed-text for semantic searchnpm install -g @djolex999/vir-cli
vir init # guided wizard: provider, models, vault, cadence,
# and an optional web-articles (raw/) folder
vir run # one pass over your sessions → notes in your vault
vir schedule install # register the daemon (runs every 3h by default)
vir init asks whether you save web articles to a folder (e.g. Obsidian Web Clipper). Point it at that raw/ directory and Vir distills those articles into the same vault. Leave it blank to keep Vir session-only.
vir schedule install works on Linux too: systemd is preferred, with cron used as a fallback when systemctl isn't available.
Vir uses TF-IDF by default. For semantic search via embeddings:
brew install ollama
ollama pull nomic-embed-text
ollama serve
Then in a new terminal:
vir embed
vir query "how do I handle rate limiting in Next.js"
Falls back to TF-IDF automatically if Ollama is not running.
Vir uses MMR (Maximum Marginal Relevance) reranking to balance relevance and diversity in query results. Instead of returning 5 notes that all say similar things, you get 5 notes covering different aspects of the topic. Tunable via retrievalDiversity in config (default 0.3, range 0.0–1.0; higher = more diverse).
Located at ~/.vir/config.json.
| Field | Default | Description | ||
|---|---|---|---|---|
vaultPath | — | Absolute path to Obsidian vault | ||
outputDir | vir | Subdir inside vault | ||
claudeProjectsDir | ~/.claude/projects | Claude Code sessions | ||
cadenceHours | 3 | Daemon run frequency (hours) | ||
provider | anthropic | anthropic or kie | ||
anthropicApiKey | — | Required if provider=anthropic | ||
kieApiKey | — | Required if provider=kie | ||
filterThreshold | 0.4 | Heuristic pre-filter (0..1) | ||
articlesDir | _(unset)_ | raw/ dir for web articles. Unset → article ingestion off | ||
distillArticles | true | Distill articles alongside sessions (needs articlesDir) | ||
filterToolCalls | moderate | Tool-output filtering: aggressive \ | moderate \ | off |
retrievalDiversity | 0.3 | MMR diversity (0..1): 0.0 = pure relevance, 1.0 = pure diversity | ||
models.classify | claude-haiku-4-5-20251001 | Classify model | ||
models.distill | claude-sonnet-4-6 | Distill model — the "smart" model for decision/large sessions | ||
models.distillFast | _(unset)_ | Cheap model for routine sessions. Set → hybrid routing on; unset → distill used for everything | ||
models.distillThreshold | 100000 | Input-token ceiling above which a session is forced to distill | ||
pricing | _(built-in)_ | Optional per-provider $/1M overrides (inputPer1M/outputPer1M). Anthropic defaults track list rates; Kie defaults are approximate — verify on your Kie dashboard |
Vir runs as an MCP server, letting Claude Code consult your vault mid-session instead of relying on static CLAUDE.md content.
Register Vir with Claude Code:
vir mcp install
Restart Claude Code. The vault is now queryable mid-session via five tools: vir_query, vir_status, vir_recent_notes, vir_recent_articles, vir_project_summary. vir_query takes a type filter (session | article | all) so Claude can scope a search to your dev sessions or your saved articles. Human-verified notes (approved via vir review) are ranked first; pass verified_only: true to vir_query or vir_recent_notes to see only those.
To unregister:
vir mcp uninstall
The LLM Wiki space has grown fast. Honest comparison:
| Vir | lucasastorian/llmwiki | Pratiyush/llm-wiki | nashsu/llm_wiki | |
|---|---|---|---|---|
| Language | TypeScript / Node | Python | Python | Cross-platform desktop |
| Distribution | npm install -g | Local app + hosted SaaS | git clone + python | Desktop app installer |
| Obsidian integration | Native (plugin v1 in dev) | Markdown output | Outputs to vault | Own UI, no Obsidian |
| Input sources | Claude Code, Web Clipper (more agents coming) | PDFs, docs upload | Claude Code, Cursor, Cline, Codex, Gemini | Documents, mixed sources |
| Retroactive on existing sessions | ✓ | n/a | from install forward | n/a |
| MCP server | ✓ | ✓ | ✓ | ✓ |
| License | MIT | open source + hosted commercial | MIT | open source |
| Vir | claude-mem | claude-memory | mem0 | |
|---|---|---|---|---|
| Reads existing Claude Code sessions | ✓ | from install forward | from install forward | n/a |
| Markdown output | ✓ | ChromaDB | LanceDB | various backends |
| Setup | npm install -g | Bun + uv + Python | pnpm + LM Studio | API/cloud setup |
| License | MIT | Apache 2.0 | MIT | open core + cloud |
高质量AI代码开发辅助工具
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。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
AI Skill Hub 点评:Claude代码维基 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 | vir |
| Topics | ai-memoryai-toolsanthropicclaude |
| GitHub | https://github.com/djolex999/vir |
| License | MIT |
| 语言 | TypeScript |
收录时间:2026-06-26 · 更新时间:2026-06-26 · License:MIT · 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