Three AI agents with distinct personalities, layered memory systems, and voice integration. Research-informed personas that learn and improve over time.
$ npm run preflight "how does authentication work"
⚡ Running preflight context search...
├── Vector search: 12 docs found
├── Graph query: 8 related nodes
└── Event log: 24 commits analyzed
✓ Context loaded: AuthContext.tsx, useAuth.ts, BUG-008
# Ready for implementation with full codebase context
EVOLUTION
The AI Workspace evolved through 9 distinct phases, from manual GPT workflows to autonomous agent orchestration.
Manual GPT workflows, no persistent context
VS Code Claude, Terminal Claude, Codex - first workspace
Event log, knowledge graph, vector store added
95-98% context reduction via tools-as-code
29 modular functions for composable workflows
Wake words, TTS, multi-agent routing with Claudia
Mika and Sloan join, native environments assigned
Cross-terminal coordination, spawn/message/close
Claudia → Nadia, Void → FZZTVLZ
PERSONAS
Each persona has unique traits, knowledge domains, and voice profiles. They collaborate on complex tasks.
Voice Concierge & Router
Complex project researcher, planner, and coordinator. Routes requests to the right function or agent.
Build Engineer & Systems Maker
Testing, validation, and systems architecture. Focused on reliability and precision.
Implementation & Coding
Deep coding, refactoring, and feature implementation. Writes production-quality code.
VOICE INTEGRATION
Wake words, push-to-talk, and persona-specific voice synthesis for natural interaction.
"Hey Nadia", "Hey Mika", "Hey Sloan" - each persona responds to their name.
Option+Space hotkey for instant voice input. Hammerspoon-powered for low latency.
Persona-specific voice profiles. Each agent sounds distinct and authentic.
OpenAI Whisper for accurate speech-to-text transcription.
# Voice detected: "Hey Mika"
→ Loading persona bundle: mika.json
→ Voice profile: ElevenLabs/mika-voice
→ Knowledge domains: agent-architecture, distributed-systems
[Mika] Ready. What do you need?
LAYERED ARCHITECTURE
Research-informed memory creates agents that improve over time. Each layer builds on the previous.
# When Mika implements a feature, she recalls:
1. Her compiled research on agent patterns
2. Knowledge of distributed systems
3. Reflections on reliability principles
4. Current session context
# This creates persona continuity across sessions
✓ Memory loaded: 4 layers, 12 knowledge files
KNOWLEDGE SYSTEMS
Vector search, knowledge graphs, and event logs create comprehensive workspace awareness.
Vector store with OpenAI embeddings. Natural language queries over entire workspace history.
Searchable archive of all agent sessions. Context continuity across conversations.
Nodes and edges mapping relationships between files, commits, bugs, and concepts.
Every action timestamped and tagged. Full audit trail for debugging and analysis.
ADVANCED CAPABILITIES
26 modular functions, multi-agent spawning, and code execution tools for complex workflows.
Context gathering before implementation. Vector + graph + events in parallel.
Parallel agent execution with inter-agent messaging. Terminal-based orchestration.
TypeScript tooling for research workflows. Automated discovery and validation.
26 modular functions for common tasks. Request classifier routes to correct handler.
# Spawn parallel agents for complex tasks
$ SPAWNING_PERSONA=nadia ./spawn-agent-session.sh sloan claude
→ Session 8642 created for sloan
$ ./send-message.sh 8642 sloan "Build portfolio page"
[sloan] On it. Loading persona bundle...
# Agents communicate via inter-terminal messaging
AUTOMATED WORKFLOWS
Headlines curation and persona content generation run as automated multi-step workflows.
run-pipeline-local.tsFetches from Spotify, YouTube, TikTok, Billboard
800+ itemscurate-headlines-manual.tsHuman-in-loop curation for hero/trending
Selected picksrun-curation.tsPopulates content cache
Curated itemsSection ComponentsHero, Trending, Charts, Videos, Fashion
7 sectionsgenerate-curation-pool.tsSources from Reddit, TikTok, Spotify, YouTube
25 itemsrun-curation-draft.tsInitializes 3-agent round-robin draft
3 personasPersona PicksEach persona picks content matching their taste
3 picks eachpersona-content-pipeline.pyGenerates authentic persona drafts
Final postsTECH STACK
The patterns here are reusable. Multi-agent orchestration, layered memory, voice integration.