YukisDB Python SDK & CLI ⚡

Official Python SDK & Developer CLI for YukisDB — Unified Cloud NoSQL JSON Document Database, 2GB Object Storage with Zero Egress Bandwidth Bills, and AI Vector Embeddings Memory.

📦 Installation

pip install yukisdb

💻 CLI Quickstart

Once installed, the yukisdb command is immediately available in your terminal:

# Check cluster telemetry & health
yukisdb health

# View developer free tier quotas
yukisdb tier

# Zero-auth sandbox insert
yukisdb sandbox insert '{"task": "test", "completed": true}'

# Zero-auth sandbox query
yukisdb sandbox find '{"completed": true}'

# Generate an ephemeral API testing key
yukisdb anonymous-key

# Authenticated CRUD (requires export YUKISDB_API_KEY=ydb_live_...)
export YUKISDB_API_KEY="ydb_live_your_key_here"
yukisdb insert users '{"name": "Sudeep", "role": "admin"}'
yukisdb find users '{"role": "admin"}'
yukisdb upload model_checkpoint.pt

🐍 Python SDK Quickstart

1. Initialize Client

from yukisdb import YukisDB

# Automatically picks up YUKISDB_API_KEY environment variable if omitted
db = YukisDB(api_key="ydb_live_your_api_key_here")

2. NoSQL Document Operations (CRUD)

# Insert a document
doc = db.collection("users").insert({
    "name": "Alice",
    "email": "alice@example.com",
    "tier": "developer"
})
print("Inserted ID:", doc["id"])

# Query documents
users = db.collection("users").find({"tier": "developer"})
print("Found users:", users)

# Retrieve single document by ID
user = db.collection("users").get(doc["id"])

# Update document
db.collection("users").update(doc["id"], {"tier": "pro"})

# Delete document
db.collection("users").delete(doc["id"])

3. AI Vector Embeddings Search

# Cosine similarity nearest-neighbor search for agent long-term memory
results = db.collection("agent_memory").vector_search(
    vector=[0.12, -0.45, 0.88, 0.05],
    field="embedding",
    top_k=5
)
for match in results:
    print(match["title"], match["_score"])

4. 2GB Object Storage (Zero Egress Bills)

# Upload any file up to 2GB with instant CDN streaming URL
file_info = db.storage.upload("video_dataset.mp4", custom_slug="dataset-2026")
print("Public CDN URL:", file_info["url"])

# List tenant storage files
files = db.storage.list()
print("Storage objects:", files)

5. Zero-Auth Public Sandbox (Prototyping)

# Store without registration
sandbox_doc = db.sandbox.insert({"experiment": "RAG pipeline test"})
print("Saved to sandbox:", sandbox_doc)

🤖 Model Context Protocol (MCP) Integration

YukisDB runs a native Model Context Protocol (MCP 2024-11-05) server for Claude Desktop, Cursor AI, and autonomous agents:

{
  "mcpServers": {
    "yukisdb": {
      "command": "yukisdb",
      "args": ["mcp"],
      "env": {
        "YUKISDB_API_KEY": "ydb_live_your_api_key_here"
      }
    }
  }
}

Or connect directly to Streamable HTTP: https://yukisdb.yukiapi.site/mcp.

🔗 Resources

• Website & Portal: https://yukisdb.yukiapi.site
• Developer Documentation: https://yukisdb.yukiapi.site/docs
• OpenAPI 3.1 Spec: https://yukisdb.yukiapi.site/openapi.json
• LLMs Knowledge Reference: https://yukisdb.yukiapi.site/llms.txt
• Support & Security: support@yukiapi.site