Docs
Third-party systems
Third-party systems
Connect Cascades workflows to SaaS platforms, operational systems, and enterprise tooling.
Cascades integrates with external systems through APIs, webhooks, and workflow tasks.
Teams commonly connect Cascades to platforms used across operations, engineering, security, and customer workflows.
Engineering systems
Examples
| Category | Examples |
|---|---|
| Issue & change tracking | Atlassian Jira, Linear, Azure DevOps Boards |
| CI/CD | GitHub Actions, GitLab CI, Jenkins, Buildkite |
| Deployment & delivery | Argo CD, Flux, Kubernetes APIs, cloud deployment APIs |
Typical patterns
- Change-linked runs — A merge or deployment event arrives via webhook; Cascades validates context (branch, environment), runs an approval or verification workflow, then posts status back to the tool or opens a follow-up task.
- Release orchestration — Coordinate gated steps (tests → staging sign-off → production) with retries and explicit failure handling instead of ad hoc scripts.
- Incident-driven remediation — Trigger workflows from engineering alerts or incident tickets (assign owners, create checklist tasks, notify channels).
See GitHub workflows and GitLab pipelines for concrete connect/webhook routes used in this repo.
Communication platforms
Examples
| Category | Examples |
|---|---|
| Chat | Slack, Microsoft Teams, Google Chat (via webhooks or bots) |
| Alerting | PagerDuty, Opsgenie, VictorOps |
| SMTP gateways, transactional email APIs (SendGrid, SES, Postmark) |
Typical patterns
- Human-in-the-loop approvals — Pause a run until an operator approves in Slack or Teams; resume execution only after the integration confirms the decision (timeout paths optional).
- Escalation ladders — On repeated failures or SLA breaches, notify successive channels (chat → page → email) using templated payloads driven by workflow outputs.
- Operational broadcasts — Summarize run outcomes (success, partial retry, permanent failure) without duplicating secrets—prefer dashboard links and masked identifiers in messages.
Incident and operational systems
Examples
| Category | Examples |
|---|---|
| Ticketing | ServiceNow, Zendesk, Freshservice, Jira Service Management |
| Monitoring & observability | Datadog, New Relic, Grafana Alerting, Prometheus Alertmanager |
| Status & collaboration | Statuspage-style APIs, shared incident docs |
Typical patterns
- Ticket-correlated runs — Create or update tickets when a workflow starts or completes; attach run IDs and proof references so responders see execution history in one place.
- Alert enrichment — Ingest monitoring webhooks, normalize severity and service tags in Cascades, then route remediation workflows or approvals based on policy.
- Post-incident tasks — Automate follow-ups (evidence collection, timeline snippets, closure criteria) as structured workflow steps rather than manual checklists only.
Business systems
Examples
| Category | Examples |
|---|---|
| CRM & sales ops | Salesforce, HubSpot, Dynamics 365 Sales |
| Billing & subscriptions | Stripe Billing, Chargebee, Zuora (often via APIs or accounting bridges) |
| Customer onboarding | Identity / profile stores, provisioning portals, entitlement catalogs |
Typical patterns
- Onboarding automation — After CRM or signup events, drive provisioning steps (internal APIs, approval gates, verification tasks) with full run visibility.
- Entitlement alignment — Reflect plan changes or seat counts by orchestrating updates across internal services while recording outcomes for audit.
- Operational workflows — Cross-functional processes (sales → finance → delivery) where Cascades sequences steps and preserves verification records where required.
Custom systems
Teams also connect:
- Internal APIs — Microservices, legacy SOAP/REST gateways, private RPC fronts exposed only inside the network.
- Proprietary services — Partner-specific endpoints, industry exchanges, or regulated data feeds authenticated per environment.
- Custom operational platforms — Internal runbooks-as-APIs, approval databases, or bespoke dashboards invoked via API tasks.
Implementation follows the same patterns as External APIs (authentication, timeouts, idempotency) and Webhooks for inbound events. For Git-centric setups, start from the Integrations reference recipes for GitHub, GitLab, and Slack.