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.
Anthropic’s AI‑Native SDLC Playbook shows how agents can accelerate coding but also expose new challenges in intent definition, specification, verification, release gating, production monitoring, and permission management, illustrated with a payment‑duplicate‑callback case and internal productivity data.
The article analyses Pi’s Harness v2 redesign, explaining how persistent transaction‑style recording, tool‑result pruning with spill, replay policies, and separated storage of conversation, runtime state, and usage enable an agent to survive process crashes and still prove which steps were completed.
Upgrading DSH to version 0.1.0‑rc.8 shows that Codex and Claude Code can be invoked as subagents, exposing the internal Harness execution system and prompting a discussion of whether agents are forming a new runtime layer in software architecture.
The article examines why ontology is essential for enterprise knowledge bases, showing how it resolves ambiguities that RAG, knowledge graphs, and agents cannot handle alone, and outlines a four‑layer architecture that ensures stable IDs, relationship semantics, fact lifecycle, and safe action execution.
A recent Agent Harness test comparing Pi, OpenCode, and DSH shows Pi completing 20 of 30 tasks at $0.028 per task versus OpenCode's 14 tasks at $0.195, while the article analyzes how each system distributes complexity, handles extensions, server boundaries, and runtime composability, highlighting trade‑offs for real‑world migrations.
With a growing ecosystem of DSH plugins—from simple file path helpers to vision toolkits and automation workflows—this article outlines a pragmatic selection order, explains the four architectural layers each plugin touches, and highlights the hidden dependencies, data persistence, permission and failure‑recovery issues you must evaluate before installing.
The article analyzes how Cordis manages plugin lifecycle issues—preventing "ghost" states after removal—by contrasting the Pi model, which delegates cleanup to extensions, with the DSH model that records cross‑plugin relationships in the runtime, detailing the mechanisms, code paths, and trade‑offs of each.
This article dissects DeepSeek Harness (DSH), showing how its plugin‑based runtime graph and event‑stream architecture let agents dynamically compose capabilities, manage lifecycles, and recover from interruptions, while comparing its design to the Pi coding agent and outlining practical presets and configuration steps.
The article examines DeepSeek’s August 13 release of V4 Pro and the new DSH runtime, breaking down the three‑layer architecture (model, Responses API protocol, and DSH runtime), benchmark scores, pricing tiers, plugin modes, session logging, and practical guidance for evaluating agent workloads and costs.
Even after a model and its API are working, real‑world AI agents often fail because of missing infrastructure such as tool definitions, sandbox boundaries, state persistence, memory handling, tracing, and evaluation, requiring a systematic approach to turn model outputs into controlled, repeatable actions.
