WisebotAI
Workflows

Templates and real use cases

Starter graphs you can rebuild in the editor—plus who uses them, how they are triggered, and links to step-by-step patterns.

Templates and real use cases

Templates here are blueprints: the same steps you add on the Build canvas, connected left → right. The product does not yet ship one-click “import template” files; copy the graph below, then tune prompts, keys, and URLs for your org.

For exact step tables and sample webhook JSON, use Example workflows after you pick a template below.


Starter templates (graphs to build)

Each row is a minimal chain. Extend it with Transform, Log, or extra LLM / HTTPS steps as needed.

TemplateGraph (in order)Typical triggersMaps to
Sanity checkInputOutputManualProves Run now and inputs work.
AI text pass-throughInput (e.g. message) → LLMOutputManual, WebhookSummarize text (any prompt, not only summaries).
Grounded answerInput (question) → KB preflightLLMOutputManual, WebhookKnowledge-assisted answer.
Classify & monitorInputLLM (strict JSON) → TransformLogOutputWebhookClassify and branch.
Notify another systemInputLLM (short summary) → HTTPS (POST) → OutputWebhookHTTP callback after enrichment.
Daily / hourly digestInput (optional defaults) → KB preflight or LLMOutputSchedule, ManualScheduled digest.
Search on a live threadInput (thread + session + query keys) → Agent tool (searchTool) → OutputWebhook, ManualAgent tool on an existing conversation.
External app actionInputMCP toolOutputWebhookMCP: create a task.
Custom scriptInputCodeOutputWebhook, ManualSandbox code.

Webhook and schedule setup live under Run & automate in the editor. Secrets and URLs are described in Triggering workflows.


Real use cases (who, trigger, outcome)

These are how teams actually use linear workflows—not a feature guarantee for every integration, but realistic designs on top of the steps above.

Support and success

Use caseWhoTriggerWhat the flow does
Ticket summary for agentsSupport leadsWebhook when a ticket is created or updatedInput receives ticket text; LLM produces bullets; Output returns summary for the helpdesk UI or a follow-up step.
Escalation prepTier-1 supportManual or webhookInput with thread/session ids; Agent tool (escalateConversationTool or your custom tool); Output returns tool result.
Canned KB reply draftSuccess / supportManualKB preflight + LLM so drafts stay close to uploaded policies; human sends the final message in the channel.

Sales and marketing

Use caseWhoTriggerWhat the flow does
Lead notes → CRM noteRevOpsWebhook from form or CRMLLM turns free-text into a short CRM-safe note; HTTPS POSTs to your API; see HTTP callback.
Inbound email classificationMarketing opsWebhookLLM returns structured JSON (priority, topic); Transform + Log for routing; pattern in Classify and branch.

Operations and internal tools

Use caseWhoTriggerWhat the flow does
Morning digestOps / leadershipScheduleSame as Scheduled digest—fixed or default Inputs plus LLM or KB preflight.
Create tracking taskAny team with MCPWebhookMCP tool to Notion, Jira, or Slack (after integration is connected)—MCP example.

Developers and platform teams

Use caseWhoTriggerWhat the flow does
CI / release blurbEngineersWebhook from CI with changelog text in JSONInputLLM (release notes style) → Output for Slack or email body.
Normalize payloadBackendWebhookInputCode (parse, validate, reshape) → HTTPS or OutputSandbox code.

Choosing a template

  1. Event from another product → start from a Webhook row, match Input keys to the JSON you will POST (see cli-and-sdk).
  2. Recurring reportSchedule + Input defaults + LLM or KB preflight.
  3. Must use live chat context → plan Input keys for thread/session and use Agent tool or Agent turn only when those ids already exist—see warnings in Example 6.