Professional architect sharing high‑quality architecture insights. Topics include high‑availability, high‑performance, high‑stability architectures, big data, machine learning, Java, system and distributed architecture, AI, and practical large‑scale architecture case studies. Open to ideas‑driven architects who enjoy sharing and learning.
In the fast‑changing Agent era, architects should focus on durable engineering capabilities—context management, tool design, evaluation, harness, permissions, and cost control—rather than chasing the latest frameworks, ensuring agents remain stable and controllable in production systems.
The article examines how a model defines an agent’s ultimate capabilities while the harness determines its production reliability, detailing continuous evaluation, context‑budgeting, tool‑error classification, multi‑model migration, and SRE‑style engineering practices needed to keep AI coding agents stable and performant.
The article examines a minimal 30‑minute Agent loop demo, then analyzes how Harness can serve as the backend by introducing a runtime capability registry, worker lifecycle management, diverse triggers, and unified tracing, outlining four concrete design actions to close the gaps for agent‑ready systems.
Karpathy’s 2026 Sequoia AI Ascent interview shows that while Vibe Coding lowers the barrier for rapid prototyping, the emerging Agentic Engineering paradigm pushes AI agents into the full software‑development lifecycle, demanding new control planes, verification, context handling and blurring the line between senior engineers and architects.
The article dissects Hermes Agent’s four‑layer memory architecture—hot memory, session search, skills, and optional Honcho—explaining how each layer’s cost and purpose differ from OpenClaw’s approach, and why careful placement of facts, history, procedures, and user models leads to more stable, cache‑aware agents.
Long Claude Code sessions get polluted when exploratory commands, logs, and temporary files share the main window, so Subagents run those steps in independent workspaces, returning only concise results and preserving the main context for decision‑making.
The article argues that an agent harness’s context window should be treated as a bounded workset rather than an ever‑growing transcript, and explains how pagination, compression, tool‑output limits, session isolation, and sub‑agent design together determine whether long‑running agents remain reliable and efficient.
The article explains how to choose between Sub‑Agent and Agent Team structures for multi‑agent systems by evaluating whether sub‑tasks share context, need isolation, compression, parallelism, or continuous collaboration, and provides practical guidelines, pitfalls, and a decision framework to avoid over‑engineering.
The article argues that building products for AI agents requires more than swapping UI pages for APIs or adding MCPs; it demands reorganizing product capabilities into actions that agents can understand, invoke, be constrained by, and audit, while addressing semantics, governance, and reliability.
The DeepSeek V4 technical report shows how a 1 million‑token context forces a redesign of attention, KV‑cache, optimizer, quantization and inference budgeting, turning long‑context capability from a costly showcase into a production‑ready feature for agents, search and Chinese professional tasks.
