Voyage AI embedding and reranking CLI integrated with MongoDB Atlas Vector Search. Use for: generating text embeddings, reranking search results, storing embeddings in Atlas, performing vector similarity search, creating vector search indexes, listing available models, comparing text similarity, bulk ingestion, interactive demos, and learning about AI concepts. Triggers: embed text, generate embeddings, vector search, rerank documents, voyage ai, semantic search, similarity search, store embeddings, atlas vector search, embedding models, cosine similarity, bulk ingest, explain embeddings.
数据来源：ClawHub。 在 ClawSkills 查看
选择你使用的 Agent
方法一：命令行安装（推荐）
推荐（无需提前安装 clawhub）
npx clawhub@latest --dir ~/.claude/skills install voyageai-skill或使用 clawhub CLI（需提前安装）
clawhub --dir ~/.claude/skills install voyageai-skill⚠️ 需要 Node.js 18+，没有 Node？请使用下方方法二直接下载 ZIP。 安装 Node.js →
方法二：手动下载安装（无需 Node）
下载 ZIP，解压后将文件夹放到以下路径，重启 Agent 即可：
安装路径
~/.claude/skills/voyageai-skill/💡解压后将文件夹放到上方路径，重启 Agent 即可生效
--- name: voyageai description: > Voyage AI embedding and reranking CLI integrated with MongoDB Atlas Vector Search. Use for: generating text embeddings, reranking search results, storing embeddings in Atlas, performing vector similarity search, creating vector search indexes, listing available models, comparing text similarity, bulk ingestion, interactive demos, and learning about AI concepts. Triggers: embed text, generate embeddings, vector search, rerank documents, voyage ai, semantic search, similarity search, store embeddings, atlas vector search, embedding models, cosine similarity, bulk ingest, explain embeddings. metadata: openclaw: emoji: "🧭" author: name: Michael Lynn url: https://mlynn.org github: mrlynn version: "1.4.0" license: MIT tags: - embeddings - vector-search - reranking - mongodb - atlas - semantic-search - rag - voyage-ai requires: bins: - vai env: - VOYAGE_API_KEY install: - id: npm kind: npm package: voyageai-cli global: true ---
Uses the vai CLI (voyageai-cli) for Voyage AI embeddings, reranking, and MongoDB Atlas Vector Search. Pure Node.js — no Python required.
npm install -g voyageai-cli| Variable | Required For | Description | |----------|-------------|-------------| | VOYAGE_API_KEY | embed, rerank, store, search, similarity, ingest, ping | Model API key from MongoDB Atlas | | MONGODB_URI | store, search, index, ingest, ping (optional) | Atlas connection string |
Get your API key: MongoDB Atlas → AI Models → Create model API key
vai embed "What is MongoDB?"
vai embed "search query" --model voyage-4-large --input-type query --dimensions 512
vai embed --file document.txt --input-type document
cat texts.txt | vai embed
vai embed "hello" --output-format arrayvai rerank --query "database performance" --documents "MongoDB is fast" "SQL is relational"
vai rerank --query "best database" --documents-file candidates.json --top-k 3vai store --db mydb --collection docs --field embedding \
--text "MongoDB Atlas is a cloud database" \
--metadata '{"source": "docs"}'
# Batch from JSONL
vai store --db mydb --collection docs --field embedding --file documents.jsonlvai search --query "cloud database" --db mydb --collection docs \
--index vector_index --field embedding
# With pre-filter
vai search --query "performance" --db mydb --collection docs \
--index vector_index --field embedding --filter '{"category": "guides"}' --limit 5vai index create --db mydb --collection docs --field embedding \
--dimensions 1024 --similarity cosine --index-name my_index
vai index list --db mydb --collection docs
vai index delete --db mydb --collection docs --index-name my_indexvai models
vai models --type embedding
vai models --type reranking
vai models --jsonvai ping
vai ping --jsonvai config set api-key "pa-your-key"
echo "pa-your-key" | vai config set api-key --stdin
vai config get
vai config delete api-key
vai config path
vai config resetvai demo
vai demo --no-pause
vai demo --skip-pipeline
vai demo --keepvai explain # List all topics
vai explain embeddings
vai explain reranking
vai explain vector-search
vai explain rag
vai explain cosine-similarity
vai explain two-stage-retrieval
vai explain input-type
vai explain models
vai explain api-keys
vai explain api-access
vai explain batch-processingvai similarity "MongoDB is a document database" "MongoDB Atlas is a cloud database"
vai similarity "database performance" --against "MongoDB is fast" "PostgreSQL is relational"
vai similarity --file1 doc1.txt --file2 doc2.txt
vai similarity "text A" "text B" --jsonvai ingest --file corpus.jsonl --db myapp --collection docs --field embedding
vai ingest --file data.csv --db myapp --collection docs --field embedding --text-column content
vai ingest --file corpus.jsonl --db myapp --collection docs --field embedding \
--model voyage-4 --batch-size 100 --input-type document
vai ingest --file corpus.jsonl --db myapp --collection docs --field embedding --dry-runvai completions bash # Output bash completion script
vai completions zsh # Output zsh completion script
# Install bash completions
vai completions bash >> ~/.bashrc && source ~/.bashrc
# Install zsh completions
vai completions zsh > ~/.zsh/completions/_vaivai help
vai help embed
vai embed --help# 1. Store documents
vai store --db myapp --collection articles --field embedding \
--text "MongoDB Atlas provides a fully managed cloud database" \
--metadata '{"title": "Atlas Overview"}'
# 2. Create index
vai index create --db myapp --collection articles --field embedding \
--dimensions 1024 --similarity cosine --index-name article_search
# 3. Search
vai search --query "how does cloud database work" \
--db myapp --collection articles --index article_search --field embedding# 1. Get candidates via vector search
vai search --query "database scaling" --db myapp --collection articles \
--index article_search --field embedding --limit 20 --json > candidates.json
# 2. Rerank for precision
vai rerank --query "database scaling" --documents-file candidates.json --top-k 5# 1. Validate data (dry run)
vai ingest --file corpus.jsonl --db myapp --collection docs --field embedding --dry-run
# 2. Ingest with progress
vai ingest --file corpus.jsonl --db myapp --collection docs --field embedding
# 3. Create index
vai index create --db myapp --collection docs --field embedding \
--dimensions 1024 --similarity cosine| Flag | Description | |------|-------------| | --json | Machine-readable JSON output | | --quiet | Suppress non-essential output |
安装 Voyage AI CLI 后，可以对 AI 说这些话来触发它
Help me get started with Voyage AI CLI
Explains what Voyage AI CLI does, walks through the setup, and runs a quick demo based on your current project
Use Voyage AI CLI to voyage AI embedding and reranking CLI integrated with MongoDB Atlas...
Invokes Voyage AI CLI with the right parameters and returns the result directly in the conversation
What can I do with Voyage AI CLI in my developer & devops workflow?
Lists the top use cases for Voyage AI CLI, with example commands for each scenario
将技能文件夹放到 ~/.claude/skills/voyageai-skill/ 目录（个人级，所有项目可用），或 .claude/skills/voyageai-skill/（项目级）。重启 AI 客户端后，用 /voyageai-skill 主动调用，或让 AI 根据上下文自动发现并使用。
Voyage AI CLI 支持 Claude、Cursor、OpenClaw，可与这些 AI 平台无缝集成，扩展其能力。
Voyage AI CLI 可免费安装使用。请查阅仓库了解许可证信息。
Voyage AI embedding and reranking CLI integrated with MongoDB Atlas Vector Search. Use for: generating text embeddings, reranking search results, storing embeddings in Atlas, performing vector similarity search, creating vector search indexes, listing available models, comparing text similarity, bulk ingestion, interactive demos, and learning about AI concepts. Triggers: embed text, generate embeddings, vector search, rerank documents, voyage ai, semantic search, similarity search, store embeddings, atlas vector search, embedding models, cosine similarity, bulk ingest, explain embeddings.
Automate my developer & devops tasks using Voyage AI CLI
Identifies repetitive steps in your workflow and sets up Voyage AI CLI to handle them automatically
Voyage AI CLI 属于「Developer & DevOps」分类，该分类的技能帮助 AI 智能体在此领域执行专业任务。