Integrations Overview
Connect external systems and services as tools agents can use
In Supervity, integrations are tools that agents are allowed to use to take action.
They enable agents to:
- read data from external systems
- write or update records
- trigger actions across tools
- coordinate work across platforms
Integrations are not scripts or static connectors.
They are governed capabilities enforced by the platform.
How Integrations Fit Into the Platform
- agents decide when a tool is needed
- workflows define how it executes
- the platform enforces security, permissions, and auditability
Types of Integrations
Built-In Integrations
Built-in integrations are maintained and supported by Supervity.
They are:
- preconfigured
- secure by default
- continuously updated
- optimized for agent execution
Examples include:
- communication and messaging tools
- file storage and document systems
- crm and business applications
- productivity tools
- dev and ops platforms
Custom Integrations (External Tools & APIs)
For systems not available out of the box, you can extend Supervity by:
- registering rest apis as tools
- connecting internal services
- wrapping legacy systems
- exposing custom business actions
Once registered, these tools behave exactly like built-in integrations.
→ learn more: Tooling & External APIs
Common Integration Categories
Communication & Messaging
- gmail, outlook
- slack, microsoft teams
- sms and notification services
typical actions
- send messages
- read incoming communication
- notify stakeholders
File Storage & Documents
- google drive
- onedrive
- dropbox
- enterprise file systems
typical actions
- upload and download files
- organize folders
- process documents
CRM & Sales Systems
- salesforce
- hubspot
typical actions
- create or update leads
- sync contacts
- route opportunities
Productivity & Data
- google sheets
- airtable
- excel-compatible services
typical actions
- read and write rows
- generate reports
- maintain datasets
Project & Dev Tools
- jira
- asana
- github
- gitlab
typical actions
- create tasks or issues
- track progress
- trigger workflows from events
Setting Up Integrations
Step 1: Connect an Integration
- go to settings → integrations
- select the service
- authenticate (oauth, api key, or service account)
- grant required permissions
Integrations are scoped to the workspace and governed centrally.
Step 2: Use Integrations in Agents & Workflows
Once connected, integrations become available:
- during agent planning
- as workflow steps
- as triggers or actions
- inside conditional logic
Agents automatically consider available integrations when planning execution.
Authentication & Security
Supported authentication methods:
- oauth 2.0 (preferred)
- api keys
- service accounts
- custom headers (advanced)
Security guarantees:
- credentials encrypted at rest
- least-privilege scopes
- token rotation support
- immediate access revocation
→ see details: Security & Privacy
Managing Integrations
View Connected Integrations
For each integration, you can inspect:
- connection status
- permissions granted
- last used timestamp
- associated workflows and agents
Reconnect or Rotate Credentials
If credentials expire:
- you receive a notification
- reconnect in settings
- no workflow or agent changes required
Disconnect Integrations
Disconnecting an integration:
- immediately revokes access
- prevents future executions
- does not delete workflows or agents
Any affected runs fail safely with clear errors.
Integration Governance
Integrations respect platform-level controls:
- role-based access
- workspace permissions
- approval gates for sensitive actions
- full execution logs
Admins can restrict which integrations agents are allowed to use.
Observability & Auditing
Every integration call is logged:
- which agent requested it
- which workflow step executed it
- inputs and outputs (redacted if sensitive)
- success or failure
- timestamp
This supports debugging, compliance, and cost tracking.
Best Practices
Design Safely
- grant minimum required permissions
- separate test and production integrations
- require human approval for destructive actions
Design Reliably
- handle rate limits explicitly
- batch requests where possible
- monitor failures early
Design for Scale
- prefer built-in integrations when available
- reuse tools across workflows
- monitor usage patterns over time
Who This Is For
business users
- connect everyday tools
- automate work without writing code
operations & it teams
- govern access
- standardize integrations
developers
- extend supervity with apis
- embed agent capabilities into systems
Where to Go Next
- see integrations in execution → Workflows Explained
- understand decision-making → Understanding Agents
- register custom tools → Tooling & External APIs
- trigger programmatically → API Reference
Integrations turn Supervity into a connected execution layer for agent-driven work.