Roadmap

Purpose

This page tracks the current implementation state of the lab and its planned evolution. Items are organized by status: implemented, in progress, planned, and future work (speculative).

Implemented

The following are currently operational:

Architecture

• Four-agent model: Orchestration Agent, Code Execution Agent, Research Agent, Media Agent
• Local-first operating model
• Separation of agent responsibilities and risk profiles

Orchestration Agent

• Development workflow (verify → design → prompt → review → integrate → commit)
• Memory management via conversation context and written documentation
• Coordination with Research Agent via task queue

Code Execution Agent

• Code generation and modification under explicit workflow constraints
• Isolation from git and production systems

Research Agent

• File-based task queue for research missions
• Mission Runner (run directory creation, lifecycle management)
• Phase DAG execution (F0 → F1 → F2/F4/Y1 → Y2 → F6)
• Mission artifacts (task.json, candidates.json, reviews.jsonl, price_analysis.json, youtube_transcripts.json, final_report.md, state.json)
• Media discovery tooling

Media Agent

• Stremio integration for streaming catalog search
• Chromecast integration for playback control
• Entertainment automation workflow

Documentation

• Lab Notes documentation baseline (this section)
• Bilingual content (English and Spanish)
• Mermaid diagrams in all major sections

In progress

• Orchestration workflow refinement and scope tuning
• Memory review and distillation practices
• Research pipeline domain expansion

Planned

Research Agent

• Semantic search and reranking
• Additional research domains (academic, technical documentation)
• PDF export for final reports
• Email delivery of reports

Orchestration Agent

• Structured job management (job IDs, status tracking, cancellation)
• Improved memory review workflow
• Standardized session reporting format

Infrastructure

• SQLite or PostgreSQL storage layer for artifacts and state
• Smart Orchestrator V2 with explicit job graph
• Multi-agent orchestration (coordinating multiple instances in parallel)
• Better notification system

Future work

The following are under consideration but not yet defined:

• Control UI — browser-based interface for monitoring and managing the Orchestration Agent
• Web UI dashboard — unified view of agent status, active missions, and artifact output
• Media Organizer — automated local media library management
• Plex/Radarr Integration Hub — connection between media discovery and local media server
• Stremio Addon Server — custom Stremio addon for lab-curated sources
• Advanced memory systems — persistent, retrievable, structured memory beyond conversation context

Notes

Items in the "Planned" section have defined scope and approach. Items in "Future work" are speculative — they may be redefined, merged, or dropped as the architecture evolves.

This page is updated as the lab develops.