AI

ai · · 9 min read

The Model Eats the Harness — But Only the Middle

Anthropic deleted 80% of Claude Code's system prompt with no measurable loss. A four-layer anatomy of the agent harness, three tests for what should have been a script, and a worked example where a well-written skill dissolves into a git hook.

claude-code agents skills context-engineering architecture
ai · · 8 min read

MCP vs CLI: The Data Behind the Debate

55K tokens vs 200. Perplexity dropped MCP. Here's what the benchmarks actually show about AI tool integration approaches.

mcp cli ai-tools architecture
ai · · 8 min read

The Scaling Law Plateau: What 76% of Researchers Are Saying About AGI

Scaling laws are hitting diminishing returns, and the AI research community is pivoting. Here's the data on where things actually stand.

scaling-laws agi ai-architecture transformers