Pipedream Workflows

Overview

Pipedream is a serverless integration and compute platform. It lets you build event-driven workflows with Node.js code steps, 1000+ pre-built app actions, and HTTP triggers — all running on serverless infrastructure.

Capabilities

• Build workflows with HTTP, cron, or app event triggers
• Write Node.js code steps with full npm access
• Use 1000+ pre-built app actions
• Store data in built-in Data Stores
• Make HTTP requests to any API
• Deploy as REST API endpoints
• Use connected accounts for OAuth apps

When to Use

Trigger phrases:

• "pipedream workflows"
• "Building event-driven automation with custom code"
• "Needing serverless compute for integration logic"
• "Wanting npm package access in automation workflows"

• Building event-driven automation with custom code
• Needing serverless compute for integration logic
• Wanting npm package access in automation workflows
• Building webhook receivers and API endpoints
• Replacing custom scripts with managed workflows

When NOT to Use

• Task is simple enough for Zapier (use Zapier for no-code)
• You need complex workflow orchestration (use n8n)
• Task is about data storage, not workflow automation
• You don't have API access to the services being integrated
• Task requires real-time processing (use streaming tools)
• You need to build a custom application (use development tools)

Pseudo Code

Implementation patterns for common use cases with this skill.

Workflow Structure

Trigger → Code Step → HTTP Request → IF → Pre-built Action → Code Step

Trigger Types

| Type | Config | |------|--------| | HTTP | $.interface.http — webhook endpoint | | Schedule | $.interface.timer — cron or interval | | App Event | New email, GitHub issue, etc. |

Code Step (Node.js)

// Access previous step data
const prevData = $.steps.trigger.event;

// Use npm packages
const axios = require('axios');
const _ = require('lodash');

// Make HTTP request
const response = await axios.get('https://api.example.com/data', {
  headers: { Authorization: `Bearer ${$.auth.api_key}` },
});

// Process data
const filtered = _.filter(response.data, item => item.active);

// Return data to next step
return {
  count: filtered.length,
  items: filtered,
};

HTTP Request Step

// Built-in HTTP request
const response = await $.http.request({
  method: 'POST',
  url: 'https://api.example.com/webhook',
  headers: {
    'Content-Type': 'application/json',
  },
  data: {
    event: $.steps.trigger.event.type,
    payload: $.steps.process.items,
  },
});

return response.data;

Data Store

// Get value
const count = await $.service.data_store.get('counter') || 0;

// Set value
await $.service.data_store.set('counter', count + 1);

// Delete
await $.service.data_store.delete('counter');

// List keys
const keys = await $.service.data_store.keys();

Connected Accounts

// Access connected account
const slack = $.app.slack;
const github = $.app.github;

// Use in API calls
const response = await axios.post('https://slack.com/api/chat.postMessage', {
  channel: '#alerts',
  text: `New event: ${$.steps.trigger.event.type}`,
}, {
  headers: { Authorization: `Bearer ${slack.$auth.oauth_access_token}` },
});

Deploy as API

// HTTP trigger creates an endpoint like:
// https://endpoint.p.punique.com/abc123

// Handle different methods
if ($.trigger.event.http.method === 'POST') {
  return { status: 'received', data: $.trigger.event.body };
}

if ($.trigger.event.http.method === 'GET') {
  return { status: 'ok', timestamp: Date.now() };
}

Error Handling in Steps

try {
  const result = await riskyOperation();
  return { success: true, data: result };
} catch (error) {
  // Return error to flow (doesn't stop workflow)
  return {
    success: false,
    error: error.message,
    step: 'process_data',
  };
}

Common Patterns

| Pattern | When to Use | |---------|------------| | HTTP Trigger → Process → Respond | API endpoint | | Schedule → Fetch → Transform → Store | ETL | | App Event → Code → Notify | Event-driven automation | | Webhook → Validate → Route | Multi-tenant webhooks | | Data Store get/set | Stateful workflows | | Code Step + npm | Complex logic with libraries |

Error Handling

| Error | Cause | Fix | |-------|-------|-----| | Step timeout | Long-running code | Optimize or use $.flow.delay() | | npm package not found | Not in allowed list | Use built-in $.http instead | | Data Store limit exceeded | Too many keys | Clean up old entries | | Auth error | Expired token | Reconnect account |

Red Flags

• Not testing workflows before deployment
• Ignoring error handling in workflows
• Missing logging and monitoring
• Not documenting workflow logic
• Ignoring rate limits and quotas

Verification

• [ ] Workflows are tested end-to-end
• [ ] Error handling is in place
• [ ] Logging and monitoring are configured
• [ ] Workflow logic is documented
• [ ] Rate limits are respected

Process

1. Analyze the task requirements
2. Apply domain expertise
3. Verify output quality

Anti-Rationalization

| Rationalization | Reality | |---|---| | "Manual is faster for one-off tasks" | One-off tasks become recurring. Automate early, save time later. | | "I will add error handling later" | You never do. Handle errors from day one. | | "Automation is overkill" | If you do it twice, automate it. If you do it daily, it is critical infrastructure. |

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: oyi77
• Source: oyi77/1ai-skills
• License: MIT
• Homepage: https://oyi77.github.io/1ai-skills

Install and usage instructions live in the source repository linked above.