Skip to main content

Social Media Pipeline Example

This example demonstrates how to build a workflow that processes sports events, generates AI commentary, and posts to Mastodon.

Overview​

A social media pipeline workflow follows this pattern:

Event (NFL Touchdown) → Prompt (Generate Commentary) → Action (Post to Mastodon)

Use Case​

When an NFL touchdown event is received from SportRadar, the workflow:

  1. Filters for touchdown events
  2. Generates engaging social media commentary using AI
  3. Posts the commentary to Mastodon

Step 1: Create the Event Entity​

Entity Configuration:

  • Name: "NFL Touchdown Event"
  • Type: Event
  • Condition Mode: Single Path

Condition Configuration:

{
  "type": "group",
  "operator": "AND",
  "conditions": [
    {
      "type": "rule",
      "field": "message.sport",
      "comparison": "equals",
      "value": "nfl"
    },
    {
      "type": "rule",
      "field": "message.eventType",
      "comparison": "equals",
      "value": "touchdown"
    }
  ]
}

This AND condition ensures only NFL touchdown events trigger the workflow.

Step 2: Create the Prompt Entity​

Entity Configuration:

  • Name: "Generate Touchdown Commentary"
  • Type: Prompt
  • Model: GPT-4 or Claude

Prompt Template:

You are a sports commentator writing for social media. Generate an exciting, engaging tweet about this touchdown:

Team: {{message.team}}
Player: {{message.player}}
Yards: {{message.yards}}
Quarter: {{message.quarter}}
Game Score: {{message.homeTeam}} {{message.homeScore}} - {{message.awayTeam}} {{message.awayScore}}

Requirements:
- Keep it under 280 characters
- Use 1-2 relevant emojis
- Be enthusiastic but not over the top
- Include the player's name and team

Respond with ONLY the tweet text, no explanations.

Step 3: Create the Action Entity​

Entity Configuration:

  • Name: "Post to Mastodon"
  • Type: Action

Script (Option 1 - Using helper function):

// Post the latest AI response directly to Mastodon
// Uses environment variables: MASTODON_URL, MASTODON_ACCESS_TOKEN
await postLatestPromptToMastodon();

print('Posted touchdown update to Mastodon');

Script (Option 2 - Manual control with error handling):

// Get the generated commentary
const commentary = await latestPromptResponse();

if (commentary) {
  try {
    // Post to Mastodon with custom handling
    const result = await postToMastodon(
      MASTODON_URL,
      MASTODON_ACCESS_TOKEN,
      commentary
    );
    
    print('Posted to Mastodon, status ID:', result.id);
    
    // Store the post ID for potential future reference
    output = {
      mastodonPostId: result.id,
      commentary: commentary,
      success: true
    };
  } catch (error) {
    print('Failed to post to Mastodon:', error.message);
    output = { success: false, error: error.message };
  }
} else {
  print('No commentary generated, skipping post');
}

Complete Workflow Data Structure​

{
  "nodes": {
    "n1": {
      "entityId": "event-entity-uuid",
      "name": "NFL Touchdown Event",
      "entityType": "event",
      "tfCondition": "Single Path",
      "presentation": {
        "type": "circle",
        "position": { "x": 100, "y": 200 }
      }
    },
    "n2": {
      "entityId": "prompt-entity-uuid",
      "name": "Generate Touchdown Commentary",
      "entityType": "prompt",
      "presentation": {
        "type": "brain",
        "position": { "x": 350, "y": 200 }
      }
    },
    "n3": {
      "entityId": "action-entity-uuid",
      "name": "Post to Mastodon",
      "entityType": "action",
      "presentation": {
        "type": "star",
        "position": { "x": 600, "y": 200 }
      }
    }
  },
  "connectors": {
    "c1": {
      "source": "n1",
      "target": "n2",
      "sourcePort": "out",
      "targetPort": "in"
    },
    "c2": {
      "source": "n2",
      "target": "n3",
      "sourcePort": "out",
      "targetPort": "in"
    }
  },
  "workflow": {
    "ref": "n1",
    "children": [
      {
        "ref": "n2",
        "children": [
          {
            "ref": "n3",
            "children": []
          }
        ]
      }
    ]
  },
  "presentation": {
    "viewport": { "x": 0, "y": 0, "zoom": 100 }
  }
}

Workflow Outcomes​

OutcomeWhen it occurs
SuccessAll entities execute without errors
ErrorA script throws an error or times out
IgnoreEvent condition doesn't match

Required Environment Variables​

Set these in the Admin Console under Settings → Variables:

VariableDescriptionExample
MASTODON_URLYour Mastodon instance URLhttps://mastodon.social
MASTODON_ACCESS_TOKENOAuth access tokenyour-access-token

Sample Input Message​

{
  "sport": "nfl",
  "eventType": "touchdown",
  "team": "Kansas City Chiefs",
  "player": "Patrick Mahomes",
  "yards": 15,
  "quarter": 3,
  "homeTeam": "Chiefs",
  "homeScore": 28,
  "awayTeam": "Raiders",
  "awayScore": 14,
  "timestamp": "2025-01-08T15:30:00Z"
}

Extending the Pipeline​

Add Error Handling Branch​

Use True/False branching on the Event to handle different scenarios:

{
  "n1": {
    "entityId": "event-uuid",
    "name": "NFL Touchdown Event",
    "entityType": "event",
    "tfCondition": "True/False",
    "presentation": { "type": "circle", "position": { "x": 100, "y": 200 } }
  },
  "n2": {
    "name": "true",
    "entityType": "logic-branch",
    "branchValue": "true",
    "presentation": { "type": "logic-branch", "position": { "x": 300, "y": 100 } }
  },
  "n3": {
    "name": "false",
    "entityType": "logic-branch",
    "branchValue": "false",
    "presentation": { "type": "logic-branch", "position": { "x": 300, "y": 300 } }
  }
}

Store for Analytics​

Add an action to store data for analytics:

// Store touchdown event in S3 for later analysis
await stmStore(
  ['nfl', 'touchdowns', message.team],
  {
    player: message.player,
    yards: message.yards,
    quarter: message.quarter,
    timestamp: message.timestamp
  },
  'touchdown-' + Date.now() + '.json'
);

print('Stored touchdown data for analytics');