What happens in this recording
A real narrated Lilith V3.recording.1 conversation, recorded from pinned source 4addc5aaf8ae925f3c7fb379994087afcbe8d1b0 using its version-specific test conversation.
Conversation XML
The complete XML script used for this recording, including its settings, system prompt and every scripted message.
XML / FULL RECORDING CONVERSATIONStructureInputsTools / C#StepsExit
<!-- Record with -Revision v3 for committed source; omit it to test working changes.
A .NET SDK is needed to compile the trusted greeting C# tool at runtime. -->
<video title="Lilith V3 - XML scenario sub-agents" version="3.0.0-alpha.2.recording.1" backend="Ollama" model="llama3.1:latest" mode="chat" showSubagents="false" voice="af_heart" progressSeconds="3" pauseSeconds="1.5" startupGreeting="true">
<systemPrompt>You are Lilith, a helpful local AI assistant. Keep replies under 60 words. When asked to run a scenario, use run_scenario with the saved scenario name and supplied input values. Ask for missing required input values. Never invent scenario execution. The application tracks subagents and can show concise completion summaries. Only your final reply is spoken.</systemPrompt>
<messages>
<message>Briefly explain how your startup greeting was created with a fresh scenario sub-agent.</message>
<message>Offer one example of another useful scenario, without running it, then say goodbye.</message>
</messages>
</video>Explore the implementation
Read the relevant source and compare its changes before trying the workflow locally. Video playback does not execute any tools in your browser.
Open the source archive ↗