API Integrations

Flow-Like provides comprehensive API integration capabilities—from raw HTTP requests to 300+ pre-built platform connectors for GitHub, Slack, Notion, Microsoft 365, Google Workspace, and more.

HTTP Requests

Make Any API Call

Make Request
├── URL: "https://api.example.com/users"
├── Method: GET
├── Headers: {"Authorization": "Bearer {token}"}
└── Body: (optional)
    │
    ▼
API Call (Fetch)
    │
    ▼
Response
├── Status: 200
├── Headers: {...}
└── Body: [{user data}]

Request Building

Build requests step by step:

Make Request
    │
    ▼
Set URL ("https://api.example.com/orders")
    │
    ▼
Set Method (POST)
    │
    ▼
Set Header ("Content-Type", "application/json")
    │
    ▼
Set Bearer Auth (api_key)
    │
    ▼
Set Struct Body ({ customer_id: 123, items: [...] })
    │
    ▼
API Call ──▶ Response

Response Handling

Parse responses based on content type:

API Call Response
    │
    ├──▶ To Struct (JSON) ──▶ Typed data object
    ├──▶ To Text ──▶ Raw string
    ├──▶ To Bytes ──▶ Binary data
    │
    ├──▶ Get Status ──▶ 200
    ├──▶ Is Success ──▶ true
    └──▶ Get Header ("Content-Type") ──▶ "application/json"

Streaming Responses

Handle large or streaming responses:

Streaming API Call
├── URL: "https://api.openai.com/v1/chat/completions"
├── stream: true
└── Handler: process each chunk

File Downloads

Download files from URLs:

HTTP Download
├── URL: "https://example.com/report.pdf"
└── Path: /downloads/report.pdf
    │
    ▼
File saved to path

Authentication Methods

Bearer Token

Set Bearer Auth (token)
    │
    ▼
Header: "Authorization: Bearer {token}"

API Key in Header

Set Header ("X-API-Key", api_key)

Basic Auth

Set Header ("Authorization", "Basic " + base64(user:pass))

OAuth 2.0

Pre-built OAuth flows for major platforms:

Google Provider (OAuth)
├── Scopes: ["drive.readonly", "gmail.send"]
└── Credentials from Secrets
    │
    ▼
Authenticated requests to Google APIs

Built-in Integrations

GitHub

Full GitHub API coverage with 35+ nodes:

Category	Nodes
Repos	List, Get, Search, Clone, Fork
Issues	List, Get, Create, Update, Search, Comment
Pull Requests	List, Get, Create, Merge, Review, Files
Files	Get Contents, Create/Update, Delete, Download
Branches	List, Get, Create, Delete
Commits	List, Get, Compare
Actions	List Workflows, Trigger, Get Runs, Cancel
Releases	List, Get, Create

Example: Create issue on error

Try
    │
    ▼
Your workflow
    │
    └── Catch
            │
            ▼
        GitHub Create Issue
        ├── repo: "myorg/myrepo"
        ├── title: "Workflow Error: {error_type}"
        └── body: "Error details: {error_message}"

Microsoft 365

50+ nodes for Microsoft Graph API:

Service	Capabilities
OneDrive	List, Upload, Download, Share, Copy, Move
SharePoint	Sites, Libraries, Files, Folders
Outlook	Send, List, Read, Reply, Forward, Folders
To Do	Lists, Tasks, Complete
Planner	Plans, Buckets, Tasks
Teams	(via Graph API)

Example: Save report to SharePoint

Generate Report (data)
    │
    ▼
Microsoft Provider (OAuth)
    │
    ▼
SharePoint Upload File
├── site_id: "contoso.sharepoint.com"
├── library: "Documents"
├── path: "/Reports/monthly.pdf"
└── content: report_pdf

Google Workspace

30+ nodes for Google APIs:

Service	Capabilities
Drive	List, Upload, Download, Delete, Share
Gmail	Send, Draft, List Labels
Sheets	Read, Write, Append
Slides	Create, Add Slides, Export
Meet	Create, Get Meetings
Calendar	Events management

Example: Upload to Drive and share

Google Provider (OAuth)
    │
    ▼
Drive Upload File
├── file: report.pdf
├── folder_id: "abc123"
└── name: "Q4 Report.pdf"
    │
    ▼
Get file_id
    │
    ▼
Drive Share
├── file_id: file_id
├── email: "team@company.com"
└── role: "reader"

Notion

Full Notion API integration:

Feature	Nodes
Databases	List, Get, Query
Pages	Get, Create, Update
Search	Search across workspace

Example: Sync data to Notion

SQL Query (get_customers)
    │
    ▼
For Each customer
    │
    ▼
Notion Create Page
├── database_id: "customers_db"
└── properties:
    ├── Name: customer.name
    ├── Email: customer.email
    └── Status: customer.status

Atlassian (Jira & Confluence)

60+ nodes for Atlassian products:

Jira:

Category	Nodes
Issues	Create, Get, Update, Delete, Search, Batch
Comments	Add, List
Transitions	Get, Execute
Sprints	List, Create, Start, Complete
Boards	List, Get, Config
Attachments	Get, Download, Delete

Confluence:

Category	Nodes
Spaces	List
Pages	Get, Create, Update, Delete
Search	Search content
Comments	Get, Add
Labels	Get, Add, Remove

Example: Auto-create Jira issues from form

HTTP Event (POST /submit-bug)
    │
    ▼
Jira Create Issue
├── project: "BUGS"
├── type: "Bug"
├── summary: request.title
├── description: request.description
└── priority: request.priority
    │
    ▼
Return { issue_key: "BUGS-123" }

Databricks

20+ nodes for data platform integration:

Category	Nodes
Clusters	List, Get, Start, Stop
Jobs	List, Get, Run, Status
SQL	Execute queries
DBFS	List, Read, Write
Unity Catalog	Catalogs, Schemas, Tables

Discord & Telegram

Messaging platform integrations:

Discord (25+ nodes):

Send/edit/delete messages
Channel management
DMs, reactions, polls
Interactive components (buttons, menus)

Telegram (60+ nodes):

Messages, media, files
Chats, forums, topics
Payments, games
Inline mode, commands

Example: Discord notification bot

Scheduled Event (every hour)
    │
    ▼
Check system metrics
    │
    ├── Alert condition? ──▶ Discord Send Message
    │                        ├── channel_id: "alerts"
    │                        └── content: "⚠️ High CPU: {value}%"
    │
    └── Normal ──▶ Done

Post updates and share content:

LinkedIn Provider (OAuth)
    │
    ▼
Share Text Post
├── text: "Excited to announce our Q4 results!"
└── visibility: "PUBLIC"

Webhooks (HTTP Events)

Receive incoming webhooks:

HTTP Event
├── Method: POST
├── Path: /webhooks/stripe
└── Outputs: body, headers
    │
    ▼
Verify Signature (headers.stripe-signature)
    │
    ▼
Branch: event.type
├── "payment_succeeded" ──▶ Process payment
├── "subscription_created" ──▶ Provision access
└── "invoice.paid" ──▶ Send receipt

Email (SMTP/IMAP)

Send Emails

SMTP Connection
├── host: "smtp.example.com"
├── port: 587
├── username: {from secrets}
└── password: {from secrets}
    │
    ▼
Send Mail
├── to: "customer@example.com"
├── subject: "Your order has shipped"
├── body: email_content
└── attachments: [tracking_pdf]

Read Emails

IMAP Connection
├── host: "imap.example.com"
└── credentials: {from secrets}
    │
    ▼
List Messages (folder: "INBOX", unread: true)
    │
    ▼
For Each message
    │
    ├── Extract data
    └── Process

AI Provider APIs

Pre-built integrations for 18+ AI providers:

Provider	Capabilities
OpenAI	GPT-4, Embeddings, Vision
Anthropic	Claude models
Google	Gemini models
Groq	Fast inference
Ollama	Local models
Cohere	Embeddings, Rerank
Mistral	Open models
Together AI	Open models
Perplexity	Search + AI
HuggingFace	Model hub
VoyageAI	Embeddings
xAI	Grok models

MCP (Model Context Protocol)

Connect to MCP servers for AI tools:

MCP Local Server
├── command: "npx"
├── args: ["-y", "@modelcontextprotocol/server-filesystem"]
└── env: { "ROOT_PATH": "/data" }

MCP HTTP Server
├── url: "https://mcp.example.com"
└── auth: Bearer token

DataFusion External Sources

Query external databases via SQL:

Create DataFusion Session
    │
    ▼
Register PostgreSQL ("prod_db", connection_string)
    │
    ▼
Register MySQL ("legacy_db", connection_string)
    │
    ▼
SQL Query:
"SELECT p.*, l.customer_name
 FROM prod_db.orders p
 JOIN legacy_db.customers l ON p.customer_id = l.id"

Supported sources:

PostgreSQL, MySQL, SQLite
ClickHouse, DuckDB, Oracle
Delta Lake, Iceberg
S3, Azure Blob, GCS

Patterns & Best Practices

1. Centralize Authentication

Create a dedicated board for each service:

Board: GitHubService
├── Variables:
│   └── provider: GitHubProvider (initialized once)
│
└── Quick Actions:
    ├── CreateIssue (title, body)
    ├── GetPullRequests (repo)
    └── TriggerWorkflow (workflow_id)

2. Handle Rate Limits

Retry
├── max_attempts: 3
├── delay: 1000ms
├── backoff: exponential
└── retry_on: [429, 503]
    │
    ▼
API Call

3. Validate Responses

API Call
    │
    ▼
Is Success?
├── Yes ──▶ To Struct ──▶ Process
└── No ──▶ Get Status ──▶ Handle error

4. Use Webhooks Over Polling

Instead of:

❌ Every 5 minutes: Check for new orders

Use:

✅ HTTP Event: Receive order webhook

5. Store Credentials Securely

Get Secret ("GITHUB_PAT")
    │
    ▼
GitHub Provider (token)

Never hardcode tokens in workflows.

6. Log API Calls

Console Log: "Calling {api_name} with {params}"
    │
    ▼
API Call
    │
    ▼
Console Log: "Response: {status} in {duration}ms"

Example: Multi-Platform Sync

Sync data across platforms:

HTTP Event (POST /customer-created)
    │
    ├──▶ Notion Create Page (customers database)
    │
    ├──▶ Jira Create Issue (onboarding task)
    │
    ├──▶ Slack Send Message (sales channel)
    │
    └──▶ HubSpot Create Contact (CRM)

Example: Automated Reporting

Generate and distribute reports:

Scheduled Event (Monday 9am)
    │
    ▼
DataFusion Query (weekly metrics)
    │
    ▼
Generate Charts (Nivo)
    │
    ▼
Render Template (report.md)
    │
    ▼
Convert to PDF
    │
    ├──▶ SharePoint Upload
    │
    ├──▶ Slack Post (with PDF)
    │
    └──▶ Email to stakeholders

FAQ

How do I add a new API?

Use the HTTP Request nodes for any REST API. Build request, set auth, call, parse response.

Can I use GraphQL?

Yes—use HTTP Request with POST method, set body to your GraphQL query.

What about SOAP APIs?

Use HTTP Request with appropriate XML body and headers.

How do I handle pagination?

Loop with While node, updating page/cursor until no more results.

Can I cache API responses?

Store results in Variables or Database, check before calling API.

Next Steps

Building Chatbots – Chat integrations
Data Pipelines – ETL with APIs
Building Internal Tools – UIs for API data
GenAI – AI API integrations