@gladiator1st/n8n-nodes-upstash

An all-in-one n8n Community Node Suite for Upstash Serverless Data & AI — providing native Vector Search (RAG) for AI Agents, Redis Caching & AI Rate-Limiting, and QStash Delayed Webhook Scheduling.

⚡ Superpowers Included

                     ┌─────────────────────────────────────────────────────┐
                     │          @gladiator1st/n8n-nodes-upstash            │
                     └──────────────────────────┬──────────────────────────┘
                                                │
                 ┌──────────────────────────────┴──────────────────────────────┐
                 ▼                                                             ▼
       🧠 Upstash Vector Store                                            ⚡ Upstash
    (AI Vector Store Companion)                                    (Consolidated Action Node)
    • LangChain RAG for AI Agents                                  • Resource: Vector (Upsert, Query, Fetch, Delete)
    • Server-Side Auto-Embeddings                                  • Resource: Redis (KV, TTL, JSON, Rate Limiter)
    • SQL-Like Metadata Filters                                    • Resource: QStash (Delayed Webhooks, Cron, Queue)

📦 Nodes in this Package

1. 🧠 Upstash Vector Store (UpstashVectorStore)

Plugs directly into n8n's AI Agent, Question & Answer Chain, and Vector Store Tool nodes:

• Server-Side Auto-Embeddings: Upstash embeds text documents automatically on the server side — no separate embedding model API key needed!
• External Embeddings Support: Optionally connect OpenAI, Cohere, or any n8n Embeddings node.
• Metadata Filters: Filter queries by category, dates, or custom attributes (e.g. status = "active").

2. ⚡ Upstash (Upstash)

A consolidated action node exposing all three Upstash serverless products via the Resource selector:

• Resource: Vector

  • Create or Update (Upsert): Auto-embed and insert raw text or float vector arrays with JSON metadata.
  • Query Similar: Search nearest vectors or text embeddings with cosine/dot-product similarity.
  • Fetch by ID: Batch retrieve stored documents and vectors.
  • Delete by ID: Delete vectors by ID.
  • Get Info & Reset: Inspect vector count, dimensions, and index metrics or reset namespace.

• Resource: Redis

  • Get Value & Set Value: Fast Key-Value storage with optional TTL expiration.
  • Delete Key(s) & Increment Counter: Delete keys or increment integer counters atomically.
  • JSON Get & JSON Set: Read/write structured data using JSONPath (e.g. $.user.role).
  • AI Agent Rate Limiter: Atomic sliding-window rate limiting to protect AI agent quotas and prevent runaway loops.

• Resource: QStash

  • Publish Message (Delayed/Queue): Send HTTP requests with human durations (e.g. 10s, 15m, 2h, 7d).
  • Reliable Retries: Automatic exponential backoff retries on destination failure.
  • Create Cron Schedule: Schedule recurring cron webhooks (e.g. 0 9 * * 1).
  • List & Delete Schedules / Cancel Messages: Manage active cron schedules and queued messages.

🚀 Installation

In n8n (Community Nodes)

1. In your n8n instance, navigate to Settings ➔ Community Nodes.
2. Click Install a community node.
3. Enter:

   @gladiator1st/n8n-nodes-upstash
   

4. Confirm terms and click Install.

⚙️ Credentials & Setup

Upstash offers a generous free tier (10,000 req/day for Redis & Vector, 500 messages/day for QStash) with no credit card required.

1. Sign up at upstash.com.
2. For Vector: Create a Vector Index ➔ Copy UPSTASH_VECTOR_REST_URL & UPSTASH_VECTOR_REST_TOKEN.
3. For Redis: Create a Redis Database ➔ Copy UPSTASH_REDIS_REST_URL & UPSTASH_REDIS_REST_TOKEN.
4. For QStash: Go to QStash tab ➔ Copy QSTASH_TOKEN.

📖 End-to-End Usage Examples

Example 1: AI Customer Support Agent with RAG & Rate Limiting

Build a zero-hallucination customer support chatbot that retrieves company policies from Upstash Vector while enforcing rate limits with Upstash Redis:

┌──────────────────────────────┐
│ When Chat Message Received   │
└──────────────┬───────────────┘
               │
               ▼
┌──────────────────────────────┐
│ Upstash (Resource: Redis)    │ ➔ Rate Limit: 10 requests / 60s per user
└──────────────┬───────────────┘
               │
               ▼
┌──────────────────────────────┐       ┌────────────────────────────┐
│ AI Agent (Google Gemini/LLM) │ ◄──── │ Vector Store Tool          │
└──────────────┬───────────────┘       └─────────────┬──────────────┘
               │                                     │
               │                                     ▼
               │                       ┌────────────────────────────┐
               │                       │ Upstash Vector Store       │ ➔ Auto-embeds query & retrieves policies
               │                       └────────────────────────────┘
               ▼
┌──────────────────────────────┐
│ Upstash (Resource: QStash)   │ ➔ Schedules delayed CRM follow-up email in 15 minutes
└──────────────────────────────┘

Step-by-Step Configuration:

1. Seed Knowledge: Add an Upstash node (Resource: Vector, Operation: Create or Update) with Vector ID: policy_refund, Input Type: Raw Text, and Text Data: All customers are eligible for a 100% refund within 30 days.. Execute once to index the text.
2. Rate Limiter: Add an Upstash node (Resource: Redis, Operation: AI Agent Rate Limiter), set Rate Limit Identifier: {{ $json.chatInput }}, Max Allowed Requests: 10, Window Duration: 60.
3. AI Agent & Tool: Add an AI Agent node, connect an LLM (e.g. OpenAI / Gemini) and a Vector Store Tool. Connect Upstash Vector Store to the Vector Store Tool.
4. Delayed Follow-Up: Add an Upstash node (Resource: QStash, Operation: Publish Message), set Destination URL: https://api.crm.com/leads, and Delay: 15m.
5. Chat: Ask "What is your refund policy?" in the chat box. The AI retrieves the exact 30-day refund policy from Upstash and answers accurately!

Example 2: Full Test Suite Workflow (examples/upstash_test_workflow.json)

This package includes a ready-to-run verification workflow covering all three Upstash services.

How to import and run:

1. Open n8n and click Add Workflow (or press Ctrl+O).
2. Click the ... menu in the top right ➔ Import from File.
3. Select examples/upstash_test_workflow.json.
4. Assign your Upstash credentials to the nodes.
5. Click Test workflow to verify Upsert, Semantic Search, Redis KV Caching, Rate Limiting, and QStash Queuing in one click!

📄 License

MIT © Muhammad Qasim