Workflows

Temporal jobs. POST /api/workflow/{name} starts a run and returns immediately (async: true). Poll or use a webhook for completion.

POST/api/workflow/leadGeneration

Auth: Bearer or x-api-key. Body always includes site_id (string UUID). Missing site_id → 400 { "code": "INVALID_SITE_ID" }.


curl -X POST "https://backend.makinari.com/api/workflow/leadGeneration" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"site_id":"YOUR_SITE_ID"}'

Success: { "success": true, ... } plus a Temporal workflowId such as lead-generation-{site_id}-{timestamp}. Failure: 500 WORKFLOW_EXECUTION_ERROR.

Catalog

All of these are POST /api/workflow/<name> with at least site_id. Extra fields are workflow-specific (do not invent maxLeads unless that route documents it).

CRM

Name	What it does
`leadGeneration`	Mine new leads for the site ICP
`leadFollowUp`	Sales follow-up on cold leads
`leadResearch`	Enrich / research a lead set
`enrichLead`	Fill missing contact fields
`leadInvalidation`	Mark bad leads
`assignLeads`	CMO assignment pass
`analyzeSite`	Scrape a URL against the ICP
`idealClientProfileMining`	Build / refresh ICP
`keyAccountGeneration`	Key-account list

Marketing

Name	What it does
`buildSegments`	Cluster leads into segments
`buildSegmentsICP`	Segments from ICP
`buildCampaigns`	Generate campaigns for a segment
`buildContent`	Copywriter draft + publish

Ops

Name	What it does
`customerSupport`	Analyze a ticket and draft a reply
`syncEmails`	Pull AgentMail and route
`dailyStandUp`	CMO daily summary
`humanIntervention`	Pause for a human
`startRobot` / `stopRobot` / `promptRobot`	Scrapybara robots
`webhook`	Inbound workflow webhook
`agentMessage`	Push a message into an agent

Not the same as Requirements (agents reason; workflows are fixed code). See Requirements vs workflows.