ai
QUERY = 'contents'; CATEGORY = 'ai' // 25 rows
An applied measurement
case, every week
Analytics and measurement, developer tools, server-side tracking, AI & LLM integrations: weekly real case analyses.
Compaction-Friendly Search Output: A Practical Playbook
There is a way to compress search output before the agent context window fills. file:line + 2 lines context format, file-cache, dedup, code examples for tool result wrappers, and measurement results.
Token Budget Arithmetic for Agent Search
An agent context window is a budget; search is a sub-budget. Where the ~15% search budget rule comes from, how it scales for different context window sizes, and how to catch overflow signals, with numerical examples.
Code Search for AI Agents: ripgrep, ast-grep, or Semantic?
ripgrep is fast, ast-grep is structural, semantic search is smart. But for an agent the right question is different: which backend in which order, under what token budget? Decision tree, academic evidence, and practical policy snippets.
Multi-Layer Defense for LLM Production Systems: From Layer 0 to Human-in-the-Loop
Effective defense against LLM failure modes cannot be single-layered. Mode-by-mode matrix and implementation patterns for prompt, output validation, agent guardrails, observability, and human-in-the-loop layers.
LLM Agentic Failure Modes: Task Drift, Reward Hacking, Alignment Faking and More
Eight LLM failure modes that surface only in agent and tool-use systems. Mechanism and defense for task drift, incorrect tool invocation, reward hacking, positional bias, mode collapse, degeneration loops, alignment faking, and version drift.
LLM Foundational Failure Modes: Hallucination, Sycophancy, Context Rot, Instruction Attenuation
Four LLM failure modes that emerge even in a single prompt. Mechanism, academic evidence, and defense patterns for hallucination and confabulation, sycophancy, context rot, and instruction attenuation.
LLM Behavioral Failure Modes: 12 Failure Patterns and the Defense Map
LLMs forget instructions in long sessions, fabricate information, and agree with you against their own answer. A pillar map of 12 systematic failure modes and the defense layers that address them.
AI Pair Programming: Coffee Debt, Gamifying Error Tracking
AI coding assistants make mistakes. Edit commands fail to match, bash scripts crash, users have to step in and correct. The point isn't getting frustrated at errors, it's logging them and extracting patterns. The Coffee Debt system does exactly that.
Why KISS Is More Relevant Than Ever in the Age of AI Agents
According to Gartner, over 40% of agentic AI projects will be canceled. Why? Unnecessary complexity. Kelly Johnson's KISS principle, defined in the 1960s, plays a more critical role in AI agent architecture than in classical software.
Why SoC Is Critical in AI Agent Architecture
Separation of Concerns is not just a software principle but the fundamental design decision that determines whether an AI agent architecture collapses or stands. Context boundaries, defense layers, and protocol separation.
What Is UCP? Google's Universal Commerce Protocol and the Transformation of the E-Commerce Ecosystem
Google's UCP protocol is moving e-commerce from a site-centric model to an agent-centric model. Its impact on conversion tracking, attribution, remarketing, and data ownership.
AI Agent Protocol Guide: MCP, A2A, UCP, AP2, A2UI, and AG-UI
Six open protocols shaping the AI agent ecosystem in 2026. What each one does, how they relate to each other, and practical use-case scenarios.
AI-Powered Codebase Audit: A Production-Grade Approach for Solo Entrepreneurs
Enterprise-quality codebase audit process using AI tools as a solo entrepreneur. Guardrail-driven AI-assisted coding, 6 audit tracks, and the Augmented CTO model.
Living Architecture Documentation for AI Coding Agents: Research, Approaches, and Tools
Codified Context, AGENTS.md ecosystem, C4 Model, Repomix, ADRs, Spec-Driven Development, Google Code Wiki. A comparative analysis of 11 different approaches to giving AI agents architectural context, backed by research.
Hybrid Search: Smart Search Architecture with FTS5 + Vector + RRF
Keyword search alone isn't enough. Neither is vector search. What happens when you combine them with Reciprocal Rank Fusion? Hybrid search architecture through the dnomia-knowledge implementation.