🔗 Scenario 2: Event-Driven Conversation Kickoff

Overview#

When the document processing job (Scenario 1) completes, automatically create a new conversation and pre-populate it with the document summary and extracted insights. The agent then generates actionable next steps based on the document.

Status: 🚧 Coming Soon
Time to complete: 20–30 minutes (when ready)
Technologies: Job Completion Events, Chat API, Conversation Context, Async Handoff
Prerequisites: Scenario 1 completed, Documents processed successfully

What You'll Learn#

✅ Subscribe to job completion events
✅ Create conversations programmatically with pre-populated context
✅ Trigger agent reasoning from job outputs
✅ Implement event-driven async workflows
✅ Chain jobs together using the Scheduler

Coming Soon#

This scenario builds on Scenario 1 and demonstrates how to:

Listen for job completion — Hook into job run completion events
Extract context — Get the processed document summary from the job run
Create a conversation — POST to /api/chat/conversations with initial context
Seed the agent — Pre-populate the conversation history with the document insight
Let the agent act — Agent reads context and generates next steps

Example flow:

Scenario 1 completes: Document processing done
    ↓
Event fires: "DocumentProcessed"
    ↓
Extract: Summary + key topics
    ↓
Create conversation with context
    ↓
Agent generates: "This invoice is overdue. Recommend: Send payment reminder."
    ↓
(Scenario 3 extends this: send notification)

Why This Matters#

Real-world: Most production systems are event-driven. Jobs complete → triggers the next step.
Learning: Shows how to compose Scenarios into a workflow.
Platform strength: OpenClaw .NET has first-class event support via job completion hooks.

Foundation Needed#

This scenario depends on Scenario 1 working correctly. When Scenario 2 is ready:

You'll have access to the IJobCompletionListener interface
Documentation will include curl examples for subscribing to events
Detailed code walkthrough of event wiring in Dependency Injection

Check Back Soon#

Dallas and the team are actively building this scenario. Follow the Real-World Demos README for status updates.

In the meantime:

Deep-dive into Scenario 1: Document Processing
Explore Scheduler documentation
Read about Conversation Context

Next Scenario#

Once Scenario 2 is ready: Scenario 3: Alert Orchestration — Add multi-channel notifications (Email, Slack, Teams) based on agent reasoning.