Building a personal AI assistant with Executor, MCP proxies and observability instead of one big agent
henriquebastos · x · 2026-10-16
Henrique Bastos shares the architecture of his personal AI assistant: keep the agent itself small and split the infrastructure around it — Pi-durable for persistence, Executor for orchestration, and MCP proxies to avoid giving the agent unrestricted access to everything. The diagram shows Gmail/Calendar/X services, exe.dev + Cloudflare cloud infra, Grafana + VictoriaMetrics observability, a Home Assistant + UniFi home lab, WAHA bridging WhatsApp, and ElevenLabs/Parakeet for voice.
He previously tried Hermes and OpenClaw but disliked putting everything inside the agent or using it as a centralized control loop; deterministic observability routines and Executor's app-based state control address that.
More from coding & agent
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23