ExampleIntermediate1 min readUpdated May 19, 2026

manualhttpnewsgdeltarticle-texttwo-step

Fetch First GDELT News Article Text

Manual workflow that fetches public GDELT news and returns the extracted text of the first article.

Download ZIP · 1.1 KB

Before You Import

The URLs in this example use a https://<your-gdelt-proxy> placeholder that you must replace with a service you host:

• Step 1 calls <your-gdelt-proxy>/doc with GDELT DOC 2.0 query parameters (query, mode=artlist, maxrecords, format=json) and expects a JSON response with an articles array where each article has a url. This matches what GDELT's public https://api.gdeltproject.org/api/v2/doc/doc endpoint returns, so your proxy can forward this call as-is.
• Step 2 calls <your-gdelt-proxy>/doc/content with a url parameter and expects a JSON response with a text field containing the extracted article body. GDELT's public API has no article-text-extraction endpoint, so your service must implement this part itself (for example, by fetching the article URL and extracting its readable text).

If you only need the article listing without full text, see Fetch GDELT Public News (Encoded String), which calls GDELT's public API directly.

Full JSON

Click to expand

{
  "description": "Manual workflow that fetches public GDELT news and returns the extracted text of the first article.",
  "tags": [
    "manual",
    "http",
    "news",
    "gdelt",
    "article-text",
    "two-step"
  ],
  "document": {
    "name": "Fetch First GDELT News Article Text",
    "enabled": true,
    "log_level": "info",
    "trigger": {
      "type": "manual"
    },
    "workflowInput": [
      {
        "name": "query",
        "description": "Search terms for GDELT public news (e.g., climate change, elections, ai policy)",
        "schema": {
          "type": "string",
          "required": true
        }
      },
      {
        "name": "max_records",
        "description": "Optional max number of news articles to consider (default: 50)",
        "schema": {
          "type": "string",
          "required": false
        }
      }
    ],
    "steps": [
      {
        "condition": {
          "expr": "true",
          "onFailure": "throw",
          "resultType": "any"
        },
        "action": {
          "actionType": "actions/network/http/request/send",
          "actions/network/http/request/send": {
            "method": "GET",
            "url": {
              "expr": "url_params(\"https://<your-gdelt-proxy>/doc\", {\"query\": query, \"mode\": \"artlist\", \"maxrecords\": coalesce(max_records, \"50\"), \"format\": \"json\"})",
              "onFailure": "throw",
              "resultType": "any"
            },
            "headers": [],
            "returnBuffer": false,
            "cache": true
          }
        }
      },
      {
        "condition": {
          "expr": "step_ok(0) && present(step_data(0, \"articles.0.url\"))",
          "onFailure": "throw",
          "resultType": "any"
        },
        "action": {
          "actionType": "actions/network/http/request/send",
          "actions/network/http/request/send": {
            "method": "GET",
            "url": {
              "expr": "url_params(\"https://<your-gdelt-proxy>/doc/content\", {\"url\": step_data(0, \"articles.0.url\")})",
              "onFailure": "throw",
              "resultType": "any"
            },
            "headers": [],
            "returnBuffer": false,
            "cache": true
          }
        }
      }
    ],
    "returnExpression": {
      "expr": "step_data(1, \"text\", \"\")",
      "onFailure": "throw",
      "resultType": "any",
      "strict": false,
      "useSimpleCel": false
    },
    "customAttributes": {}
  }
}

Import any example into AutoTalk5 steps · ~2 min

1. Download the ZIP using the Download ZIP button in the action card.
2. In your dashboard, go to Company Settings → Data Transfer.
3. Click Import and select the downloaded ZIP file.
4. Review the import summary and confirm.
5. Document IDs are remapped automatically — no manual editing required.

More examples

manual · assistant

Ask Assistant

Workflow that sends a message to an AI assistant and returns its response

Intermediate · 1 minmanual · http

News Search (Simplified)

Before/after comparison: how new CEL functions simplify workflow expressions. This is a manual workflow that fetches news, checks results, and returns a summary.

Intermediate · 1 minworkflow · manual

Discord Image Spam LLM Classifier

Manual workflow used by the Discord spam moderator to classify suspicious Discord images, including possible compromised older accounts, and return a compact JSON moderation verdict.

Intermediate · 1 minhook · after-create

New Contact Hook

Hook-triggered workflow that fires after a contact is created

Intermediate · 1 min