v6.1 Roadmap — which feature should we build next?

[theihtisham]

Shadow Brain v6.0 ships 22 features. Here's what's on the table for v6.1 — vote with 👍 or propose new ones in the comments.

Candidates

🧠 Cognition

• Reasoning loops (o1-style reflection) — multi-pass chain-of-thought with self-critique
• Graph-RAG — semantic retrieval across the knowledge-graph module
• Synthetic data generator — turn past memories into fine-tuning JSONL

🌐 Execution

• Browser computer-use integration — call Browser Use / Skyvern from brain
• E2B-style code sandbox — safe code execution with memory-aware context
• Real-time voice pipeline — STT → LLM → TTS (Pipecat / LiveKit)

🧪 Quality

• Automated evaluation harness — LLM-as-judge, Brier-calibrated
• Benchmark suite — reproducible comparisons

📦 Ecosystem

• VS Code extension with live brain panel
• One-click Docker deployment for team brain server
• Chrome/Edge browser extension for web-context capture

🔐 Privacy

• Homomorphic query — answer semantic queries without decrypting memories
• Differential privacy for collective learning mode

How to vote

• 👍 on the ones you want
• 💬 add your own idea as a comment
• 🚀 if it's a must-have for you

Most-wanted in the top 3 get prioritized for v6.1.