Documentation Index
Fetch the complete documentation index at: https://www.aidonow.com/llms.txt
Use this file to discover all available pages before exploring further.
Articles in this section examine AI tools as engineering infrastructure — what they reliably produce, where they fail, and the governance mechanisms required to make AI output safe for production. Includes both technical implementation guides and evidence-based assessments.
Practical Guides
Building a Billing Dashboard with AI
The prompting strategy, review protocol, and verification checkpoints that produced a production-grade billing dashboard — and the failure modes encountered when those checkpoints were skipped.
Claude Code Hooks: Hard Enforcement
Why system prompt instructions degrade under task complexity and how PreToolUse hook scripts provide session-boundary-independent enforcement of architectural constraints.
Prompt Library
A structured taxonomy of prompts organized by task type — with analysis of which prompt structures produce reliable output and which introduce variance.
AI Tool Comparison
A comparative assessment of AI coding tools across dimensions of accuracy, context retention, instruction-following, and failure recovery.
AI Limitations and Boundaries
The categories of task where current AI coding assistants consistently underperform — and the detection signals that indicate when human intervention is required.
MCP Tool Routing: Nine Servers as an Agent Operating System
How routing 91 agent tools across nine domain-scoped MCP servers — with a capability-first pre-implementation gate — prevents reinvention, reduces context bloat, and enforces organizational standards at the tool layer.
Whisper MCP Federation: Per-Actor Token Scoping in Multi-Agent Systems
How scoping each AI agent’s activity-feed token to a single actor identity makes the audit trail tamper-evident — with graceful downgrade, a four-layer token storage model, and a Redmine hook bridge that fires lifecycle events without modifying business logic.
Insights & Debate
AI Code Review Blind Spots
The systematic gaps in AI-generated code review — the security, concurrency, and cross-service integration issues that current models consistently fail to surface.
When AI Excels
The task categories where AI assistance produces the highest return: pattern replication, test generation, documentation, and refactoring of well-specified code.
When AI Fails: Cascading Errors
How AI hallucinations propagate through a codebase when review gates are absent — the error cascade pattern and the checkpoints that break it.
When AI Was Right
Case analysis of decisions where AI-generated recommendations proved correct against human skepticism — and what distinguishes those cases from false positives.
The AI Productivity Myth
A critical examination of AI productivity claims: where the gains are real, where they are measurement artifacts, and what the difference means for team planning.
Labeling AI Code
The organizational case for distinguishing AI-generated from human-authored code in version control — and the practical annotation conventions that make this tractable.
What 11 Weeks Actually Changed
A longitudinal assessment of AI-assisted development after 11 weeks of production use — the capability gains, the persistent limitations, and the workflow adjustments that proved durable.