Measuring the Agent Surface
Why codified rules get ignored, what a 283,000-character boot surface looks like, and how to audit agent instructions instead of adding more of them.
Resources
Books, tools, and news that we find useful or interesting.
Filtered by:
Clear all filters#AI(remove)
Local LLMs on CPU
A case study on engineering reliable agentic workflows on constrained CPU hardware using DSPy and asynchronous architecture.
LLM Predictions for 2026
Simon Willison on the rise of coding agents, the coming security 'Challenger disaster', and the Jevons paradox.
Building effective AI agents with MCP
Red Hat explains the Model Context Protocol (MCP) as the standardized interface for enterprise AI agents.
AI & Humans: Making the Relationship Work
A reality check on how we live and work with AI agents by Bruce Schneier and Nathan E. Sanders.
DSPy: Programming (not prompting) LMs
A framework from Stanford that changes how we work with AI by defining logic with code and metrics.