Tool output not visible to agent (native agent node + verified community node)

Describe the problem/error/question

Short: Tool output never “reaches” the agent.

Long:
There appears to be a discrepancy when using a node manually versus as a tool:

• When I utilize a specific node manually, it returns the expected values.
• However, when using the same node as a tool (and the AI provides the exact same input as I do manually), the results are not being returned to the agent.

The issue is clearly visible in the “last output” (see below); the “observation” field is empty. I’ve attempted further investigation but haven’t been able to expose any relevant data. It seems the complex JSON returned by the tool is simply not reaching the agent.

Screenshots:
Overview:

Agent log:

Tool log:

Manual usage DOES work as expected:

The issue appears to be identical to the one discussed in this (closed) topic:

What is the error message (if any)?

None; the agent “thinks” all operations are successful.

Please share your workflow

{
"nodes": [
{
"parameters": {},
"type": "n8n-nodes-base.manualTrigger",
"typeVersion": 1,
"position": [
0,
0
],
"id": "b9e3978d-3d6f-4638-9b92-c85b4ef23ef5",
"name": "When clicking ‘Execute workflow’"
},
{
"parameters": {
"promptType": "define",
"text": "Search for the valid term-value of "hubspot" using your tool.",
"options": {
"returnIntermediateSteps": true
}
},
"type": "@n8n/n8n-nodes-langchain.agent",
"typeVersion": 2.1,
"position": [
272,
0
],
"id": "b14b01af-51a7-45f4-b07f-b7ea22ec4ffe",
"name": "AI Agent"
},
{
"parameters": {
"model": {
"_rl": true,
"value": "gpt-4.1",
"mode": "list",
"cachedResultName": "gpt-4.1"
},
"options": {
"maxTokens": 4000
}
},
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"typeVersion": 1.2,
"position": [
144,
208
],
"id": "bf5cae14-6b4c-4299-9f6e-4fe697b7b5e8",
"name": "OpenAI Chat Model",
"credentials": {
"openAiApi": {
"id": "EqrM5wAZrhmSzbIH",
"name": "OpenAi account"
}
}
},
{
"parameters": {
"sessionIdType": "customKey",
"sessionKey": "={{ $execution.mode === ‘test’ ? ‘$workflow.id’ + ‘119’ : $execution.id }}"
},
"type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
"typeVersion": 1.3,
"position": [
304,
208
],
"id": "6659342a-ab5f-406b-bc17-2a6c8a97ab3a",
"name": "Simple Memory"
},
{
"parameters": {
"operation": "getsuggest",
"suggestiontype": "apps",
"q": "hubspot",
"requestOptions": {}
},
"type": "@bedrijfsdatanl/n8n-nodes-bedrijfsdata.bedrijfsdata",
"typeVersion": 1,
"position": [
464,
400
],
"id": "328d1062-dc5e-4395-b7ca-b81b0e8671bf",
"name": "Get parameter suggestions",
"credentials": {
"bedrijfsdataApi": {
"id": "G5an8JqsPZzMrxn7",
"name": "TEMP"
}
}
},
{
"parameters": {
"descriptionType": "manual",
"toolDescription": "Get correct spelling of a specific app to use as a value elsewhere. This tool auto-completes input.",
"operation": "getsuggest",
"suggestiontype": "apps",
"q": "={{ /n8n-auto-generated-fromAI-override/ $fromAI(‘SearchQuery’, ``, ‘string’) }}",
"requestOptions": {}
},
"type": "@bedrijfsdatanl/n8n-nodes-bedrijfsdata.bedrijfsdataTool",
"typeVersion": 1,
"position": [
464,
192
],
"id": "01549bcb-f782-4ad5-9438-62282799c0ab",
"name": "Get parameter suggestions in Bedrijfsdata",
"credentials": {
"bedrijfsdataApi": {
"id": "G5an8JqsPZzMrxn7",
"name": "TEMP"
}
}
}
],
"connections": {
"When clicking ‘Execute workflow’": {
"main": [
[
{
"node": "AI Agent",
"type": "main",
"index": 0
}
]
]
},
"OpenAI Chat Model": {
"ailanguageModel": [
[
{
"node": "AI Agent",
"type": "ailanguageModel",
"index": 0
}
]
]
},
"Simple Memory": {
"aimemory": [
[
{
"node": "AI Agent",
"type": "aimemory",
"index": 0
}
]
]
},
"Get parameter suggestions in Bedrijfsdata": {
"aitool": [
[
{
"node": "AI Agent",
"type": "aitool",
"index": 0
}
]
]
}
},
"pinData": {},
"meta": {
"templateCredsSetupCompleted": true,
"instanceId": "04f23e4c433f478fbd7cc8c4ceff447b7061f1e9f3641a5ebf456657ac823807"
}
}

Share the output returned by the last node

[
{
"output": "It appears that there were no valid term-values found for "hubspot" in the database. This may suggest a spelling variation, absence in the source, or a temporary issue. If you have an alternative spelling or a similar term to search for, please let me know!",
"intermediateSteps": [
{
"action": {
"tool": "GetparametersuggestionsinBedrijfsdata",
"toolInput": {
"SearchQuery": "hubspot"
},
"toolCallId": "callR74WsDlokcKe053esDm2OSqG",
"log": "Invoking "GetparametersuggestionsinBedrijfsdata" with {"SearchQuery":"hubspot"}n",
"messageLog": [
{
"lc": 1,
"type": "constructor",
"id": [
"langchaincore",
"messages",
"AIMessageChunk"
],
"kwargs": {
"content": "",
"additionalkwargs": {
"toolcalls": [
{
"index": 0,
"id": "callR74WsDlokcKe053esDm2OSqG",
"type": "function",
"function": {
"name": "GetparametersuggestionsinBedrijfsdata",
"arguments": "{"SearchQuery":"hubspot"}"
}
}
]
},
"responsemetadata": {
"prompt": 0,
"completion": 0,
"usage": {
"prompttokens": 484,
"completiontokens": 24,
"totaltokens": 508,
"prompttokensdetails": {
"cachedtokens": 0,
"audiotokens": 0
},
"completiontokensdetails": {
"reasoningtokens": 0,
"audiotokens": 0,
"acceptedpredictiontokens": 0,
"rejectedpredictiontokens": 0
}
},
"finishreason": "toolcalls",
"systemfingerprint": "fp51e1070cf2",
"modelname": "gpt-4.1-2025-04-14",
"servicetier": "default"
},
"toolcallchunks": [
{
"name": "GetparametersuggestionsinBedrijfsdata",
"args": "{"SearchQuery":"hubspot"}",
"id": "callR74WsDlokcKe053esDm2OSqG",
"index": 0,
"type": "toolcallchunk"
}
],
"id": "chatcmpl-BwR31HsrTKYUpQ2AX3Lvbrp2Bcmsx",
"usagemetadata": {
"inputtokens": 484,
"outputtokens": 24,
"totaltokens": 508,
"inputtokendetails": {
"audio": 0,
"cacheread": 0
},
"outputtokendetails": {
"audio": 0,
"reasoning": 0
}
},
"toolcalls": [
{
"name": "GetparametersuggestionsinBedrijfsdata",
"args": {
"SearchQuery": "hubspot"
},
"id": "callR74WsDlokcKe053esDm2OSqG",
"type": "toolcall"
}
],
"invalidtool_calls": null
}
}
],
"observation": ""
}
}
]
}
]

Information on your n8n setup

• n8n cloud
• n8n version: 1.102.4
• Database: default
• n8n EXECUTIONS_PROCESS setting: default
• Running n8n via n8n cloud