Unrecognized node type: n8n-nodes-searxng.searxngTool

[sivom]

Installed through webui and followed the installation instructions from the below link
https://github.com/sammcj/n8n-nodes-searxng

Below is the error log

Community package installed: n8n-nodes-searxng
Unrecognized node type: n8n-nodes-searxng.searxngTool
Error: Unrecognized node type: n8n-nodes-searxng.searxngTool
    at PackageDirectoryLoader.getNode (/usr/local/lib/node_modules/n8n/node_modules/n8n-core/dist/nodes-loader/directory-loader.js:156:15)
    at LoadNodesAndCredentials.getNode (/usr/local/lib/node_modules/n8n/dist/load-nodes-and-credentials.js:277:23)
    at NodeTypes.getByNameAndVersion (/usr/local/lib/node_modules/n8n/dist/node-types.js:36:51)
    at new Workflow (/usr/local/lib/node_modules/n8n/node_modules/n8n-workflow/dist/Workflow.js:56:39)
    at TestWebhooks.toWorkflow (/usr/local/lib/node_modules/n8n/dist/webhooks/test-webhooks.js:290:16)
    at TestWebhooks.needsWebhook (/usr/local/lib/node_modules/n8n/dist/webhooks/test-webhooks.js:169:31)
    at WorkflowExecutionService.executeManually (/usr/local/lib/node_modules/n8n/dist/workflows/workflow-execution.service.js:114:58)
    at processTicksAndRejections (node:internal/process/task_queues:95:5)
    at WorkflowsController.runManually (/usr/local/lib/node_modules/n8n/dist/workflows/workflows.controller.js:293:16)
    at handler (/usr/local/lib/node_modules/n8n/dist/decorators/controller.registry.js:93:24)

[ihor-sokoliuk]

@sivom I had the same issue.
Here is the original ticket with a solution:
n8n-io/n8n#12593
Another link with my solution that I use everyday:
drudge/n8n-nodes-puppeteer#59 (comment)

[LasyShark]

@sivom Did you get this issue resolved that you could comment on? Any alternatives?
@ihor-sokoliuk No complete solution is readily apparent. Could you expand more on the following?

In my case, after this in my Dockerfile:
USER root
RUN sed -i "s/nodeType.startsWith('n8n-nodes-base') && //g" /usr/local/lib/node_modules/n8n/dist/node-types.js
USER node

Question: should we add the three lines at the bottom of the docker container code?

[ihor-sokoliuk]

@LasyShark this is not something that you can fix within the node source code
It's an n8n limitations, check the original issue thread.

I am running my n8n in docker so I have my custom n8n docker image that I build with the Dockerfile.
This Dockerfile has these lines to turn the limitations off

[rjbaat]

@ihor-sokoliuk I would like to test your searxng node and try to run a custom build n8n container.
Can you point me to the Dockerfile you adjusted in the forked n8n. I can't find where to put the snippet you mentioned above.

USER root
RUN sed -i "s/nodeType.startsWith('n8n-nodes-base') && //g" /usr/local/lib/node_modules/n8n/dist/node-types.js
USER node

[ihor-sokoliuk]

@rjbaat I do not fork from the n8n repo.
I write my own Dockerfile with the n8n as a base image.

My Dockerfile looks like this:

FROM n8nio/n8n:next # can be any latest version

USER root

# Remove limitation on the custom nodes as a tool
RUN sed -i "s/nodeType.startsWith('n8n-nodes-base') && //g" /usr/local/lib/node_modules/n8n/dist/node-types.js

# Install NPM libs that I need
....

USER node

And then build and run the image based on this Dockerfile.

I prefer docker compose file:

services:
  n8n:
    build:
      context: .
      dockerfile: Dockerfile
    restart: unless-stopped
    container_name: n8n
    ports:
      - "443:443"
    volumes:
      - ...

That builds the n8n image with your changes and will unblock the community tool usage in AI agent.

[netroy]

starting the last release, you don't need to do any of this. You can simply set the env variable N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE to true to allow community nodes as tools.

[rjbaat]

@netroy oke cool, from what version is this ENV supported? I tried 1.77.3 but I get this error now:

Unrecognized node type: n8n-nodes-searxng.searxngTool

[netroy]

never mind. I forgot that this was merged only 6 days ago, and it's yet to be a part of a release. If you want to test it, you could use the latest nightly release.

[rjbaat]

Oke thnx! I tried with nightly and I don't get the error now. Only I get another error.

Maybe @ihor-sokoliuk you can help me out?

I am running the searing container next to n8n and connected them to the same network.
Also I added this to the query:

{
  "query": "your search terms",
  "categories": ["general"],
  "additionalFields": {
    "language": "en",
    "time_range": "month",
    "safesearch": "1"
  }
}

But I get a 403 error from searxng.
Also I tried with the expression: {{ $fromAI('search_query') }}. but that gives me a 403 aswell.

I used the default API Key from settings.yaml and also tried without key.

Do you have an idea what the problem could be?

[ihor-sokoliuk]

@rjbaat you can share your node config by clicking on the searxng node at the n8n web ui, copying it (Ctrl+C) and pasting here in the code block.
It will paste the JSON representation of the n8n node.

For example, here is my searxng node as a tool that works fine:

{
  "nodes": [
    {
      "parameters": {
        "query": "={{ $fromAI(\"search_query\") }}",
        "additionalFields": {}
      },
      "type": "n8n-nodes-searxng.searxngTool",
      "typeVersion": 1,
      "position": [
        760,
        0
      ],
      "id": "57275934-4750-4161-afd7-0b448e35470c",
      "name": "Searxng",
      "credentials": {
        "searxngApi": {
          "id": "SjgKUkKRe8iOc9S8",
          "name": "Searxng account"
        }
      }
    }
  ],
  "connections": {
    "Searxng": {
      "ai_tool": [
        []
      ]
    }
  },
  "pinData": {},
  "meta": {
    "templateCredsSetupCompleted": true,
    "instanceId": "32e0122e4e877b910e74d970ea713024c39bbd223514673968296aba6460ef20"
  }
}

There could be an issue with the node setup.

Also, please consider enabling the JSON format for the SearXNG server:
open-webui/open-webui#2824

For example, this is the settings.yml for my SearXNG self-hosted server.

# see https://docs.searxng.org/admin/settings/settings.html#settings-use-default-settings
use_default_settings: true
server:
...
  limiter: false  # can be disabled for a private instance
...
ui:
  static_use_hash: true
redis:
  url: redis://redis:6379/0
general:
  default_lang: en
  enable_metrics: false
...
search:
  safe_search: 0
  autocomplete: ""
...
  formats:
    - json
    - html