I'll use the Workflow tool to orchestrate this multi-agent blog post writing task.
{
"agents": [
{
"name": "content-researcher",
"task": "Extract key technical decisions, file structures, and architectural choices from the session transcript and file list. Identify: (1) exact module names and paths used in Dablio, (2) the tier verification system (Tier 1-6), (3) free/open-source tool choices vs paid alternatives, (4) safety/audit mechanisms, (5) persistent memory architecture."
},
{
"name": "blog-writer",
"task": "Write a 600-1200 word technical blog post in HTML format covering: What Was Done (Dablio build, free voice stack decision), Technical Details (tier architecture, module breakdown, STT/TTS choices), Infrastructure (config.toml, workspace layout), Key Decisions (whisper.cpp vs Deepgram, native say vs ElevenLabs, gating/audit), What's Next. Use /
/////- tags. Target audience: Sergio and engineers. Be specific: file paths (dablio/brain.py, dablio/voice/stt.py, decisions/0001-free-voice-stack.md), function names, config keys. Explain WHY (cost, offline capability, integration simplicity). Output only HTML body content."
}
],
"dependencies": [
{
"from": "content-researcher",
"to": "blog-writer",
"key": "technical_summary"
}
]
}
///- tags. Target audience: Sergio and engineers. Be specific: file paths (dablio/brain.py, dablio/voice/stt.py, decisions/0001-free-voice-stack.md), function names, config keys. Explain WHY (cost, offline capability, integration simplicity). Output only HTML body content."
}
],
"dependencies": [
{
"from": "content-researcher",
"to": "blog-writer",
"key": "technical_summary"
}
]
}