n8n workflow template

Knowledge Base Search Sub-workflow

Called as a tool by other workflows (e.g. the WhatsApp agent). Embeds the incoming query with OpenAI, performs a vector similarity search against Supabase, and returns a numbered context block plus a source list for citations.

Starter10-15 minutesn8nOpenAISupabase

What this workflow handles

Drop-in RAG retrieval tool for any n8n AI agent
Source-attributed answers

Setup steps

1Connect OpenAI and Supabase credentials
2Point the vector search request at your Supabase RPC/table endpoint
3Call this workflow as a tool from your main agent workflow

Values to replace

Review every placeholder below before activating this workflow. Public downloads should never include live credentials or client data.

Supabase vector table endpoint URL
OpenAI credential

Workflow JSON preview

Review before downloading or importing.

Download JSON Open raw

{
  "name": "search_kb",
  "nodes": [
    {
      "parameters": {
        "workflowInputs": {
          "values": [
            {
              "name": "user_query"
            }
          ]
        }
      },
      "id": "c055762a-8fe7-4141-a639-df2372f30060",
      "typeVersion": 1.1,
      "name": "When Executed by Another Workflow",
      "type": "n8n-nodes-base.executeWorkflowTrigger",
      "position": [
        464,
        480
      ]
    },
    {
      "parameters": {
        "method": "POST",
        "url": "https://api.openai.com/v1/embeddings",
        "authentication": "predefinedCredentialType",
        "nodeCredentialType": "openAiApi",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={\n  \"model\": \"text-embedding-3-small\",\n  \"input\": \"{{ $json.user_query }}\"\n}",
        "options": {}
      },
      "id": "076080af-c0e3-4ce1-b868-eb33255a1cd7",
      "name": "Embed Question",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.4,
      "position": [
        720,
        480
      ],
      "retryOnFail": true,
      "credentials": {
        "openAiApi": {
          "id": "YOUR_CREDENTIAL_ID",
          "name": "OpenAI account"
        }
      }
    },
    {
      "parameters": {
        "method": "POST",
        "url": "LINK TO YOUR SUPABASE VECTOR TABLE",
        "authentication": "predefinedCredentialType",
        "nodeCredentialType": "supabaseApi",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={\n  \"query_embedding\": {{ JSON.stringify($json.data[0].embedding) }},\n  \"match_count\": 5\n}",
        "options": {}
      },
      "id": "4afd33e7-92d0-43b0-ad05-d7eb35882000",
      "name": "Supabase: Vector Search",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.4,
      "position": [
        960,
        480
      ],
      "retryOnFail": true,
      "credentials": {
        "supabaseApi": {
          "id": "YOUR_CREDENTIAL_ID",
          "name": "Supabase account"
        }
      }
    },
    {
      "parameters": {
        "jsCode": "// Build the context block (numbered chunks) and a matching sources list for citations.\nconst question = $('When Executed by Another Workflow').first().json.user_query|| $('Ask Question (Webhook)').item.json.question\n  || '';\n\nconst matches = $input.all().map(i => i.json);\n\nconst contextParts = matches.map((m, idx) => `[${idx + 1}] ${m.content}`);\nconst context = contextParts.join('\\n\\n');\n\nconst sources = matches.map((m, idx) => ({\n  n: idx + 1,\n  title: m.metadata?.title || 'Untitled',\n  url: m.metadata?.notion_url || ''\n}));\n\nreturn [{ json: { question, context, sources } }];"
      },
      "id": "531471b9-172b-4c5f-9dac-ac6928fb1af8",
      "name": "Build Context & Sources",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        1184,
        480
      ]
    }
  ],
  "pinData": {},
  "connections": {
    "When Executed by Another Workflow": {
      "main": [
        [
          {
            "node": "Embed Question",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Embed Question": {
      "main": [
        [
          {
            "node": "Supabase: Vector Search",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Supabase: Vector Search": {
      "main": [
        [
          {
            "node": "Build Context & Sources",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Build Context & Sources": {
      "main": [
        []
      ]
    }
  },
  "active": false,
  "settings": {
    "executionOrder": "v1",
    "binaryMode": "separate",
    "timeSavedMode": "fixed",
    "saveDataErrorExecution": "all",
    "saveManualExecutions": true,
    "callerPolicy": "workflowsFromSameOwner",
    "availableInMCP": false
  },
  "versionId": "47df270b-9bf4-4432-8dc2-eb8cdc021ffe",
  "meta": {
    "templateCredsSetupCompleted": true
  },
  "nodeGroups": [],
  "id": "AvvHbJ6ltlTEfPPa",
  "tags": []
}